In this paper. we develop new methods to automatically detect the onset and duration of freezing of gait (FOG) in people with Parkinson disease (PD) in real time, using inertial sensors. We first build a physical model that describes the trembling motion during the FOG events. Then, we design a generalized likelihood ratio test framework to develop a two-stage detector for determining the zero-velocity and trembling events during gait. Thereafter, to filter out falsely detected FOG events, we develop a point-process filter that combines the output of the detectors with information about the speed of the foot, provided by a foot-mounted inertial navigation system. We computed the probability of FOG by using the point-process filter to determine the onset and duration of the FOG event. Finally, we validate the performance of the proposed system design using real data obtained from people with PD who performed a set of gait tasks. We compare our FOG detection results with an existing method that only uses accelerometer data. The results indicate that our method yields 81.03% accuracy in detecting FOG events and a threefold decrease in the false-alarm rate relative to the existing method.

The problem of path planning for mobilesensors with the task of target monitoring is considered. A receding horizon optimal control approach based on the information filter is presented, where the limited field of view of the sensor can be modeled by introducing binary variables. The resulting nonlinear mixed integer problem to be solved in each sample, with no apparent tractable solution, is shown to be equivalent to a problem that robustly can be solved to global optimality using off-the-shelf optimization tools.

The kernel-based regularization method has two core issues: kernel design and hyperparameter estimation. In this paper, we focus on the second issue and study the properties of several hyperparameter estimators including the empirical Bayes (EB) estimator, two Steins unbiased risk estimators (SURE) (one related to impulse response reconstruction and the other related to output prediction) and their corresponding Oracle counterparts, with an emphasis on the asymptotic properties of these hyperparameter estimators. To this goal, we first derive and then rewrite the first order optimality conditions of these hyperparameter estimators, leading to several insights on these hyperparameter estimators. Then we show that as the number of data goes to infinity, the two SUREs converge to the best hyperparameter minimizing the corresponding mean square error, respectively, while the more widely used EB estimator converges to another best hyperparameter minimizing the expectation of the EB estimation criterion. This indicates that the two SUREs are asymptotically optimal in the corresponding MSE senses but the EB estimator is not. Surprisingly, the convergence rate of two SUREs is slower than that of the EB estimator, and moreover, unlike the two SUREs, the EB estimator is independent of the convergence rate of Phi(T)Phi/N to its limit, where Phi is the regression matrix and N is the number of data. A Monte Carlo simulation is provided to demonstrate the theoretical results. (C) 2018 Elsevier Ltd. All rights reserved.

There are two primary sources of sensor measurements for driver behavior profiling within insurance telematics and fleet management. The first is the on-board diagnostics system, typically found within most modern cars. The second is the global navigation satellite system, whose associated receivers commonly are embedded into smartphones or off-the-shelf telematics devices. In this paper, we present maximum likelihood and maximum a posteriori estimators for the problem of fusing speed measurements from these two sources to jointly estimate a vehicles speed and the scale factor of the wheel speed sensors. In addition, we analyze the performance of the estimators by use of the Cramer-Rao bound, and discuss the estimation of model parameters describing measurement errors and vehicle dynamics. Last, simulations and real-world data are used to show that the proposed estimators yield a substantial performance gain compared to when employing only one of the two measurement sources.

Introduction: The use of left ventricular assist device (LVAD) has grown rapidly. Adverse events do continue to occur. In recent years, analysis of LVAD sound recordings emerged as a means to monitor pump function and detect pump thrombosis. The aim of this study was to characterize the sounds from HeartMate II and to evaluate the use of handheld iOS devices for sound recordings. Method: Signal analysis of LVAD sound recordings, with dedicated recording equipment and iOS devices, was performed. Two LVADs running in mock loop circuits were compared to an implanted LVAD. Spectral analysis and parametric signal models were explored to quantify the sound and potentially detect changes in it. Results: The sound recordings of two LVADs in individual mock loop circuits and a third one implanted in a patient appeared to be similar. Qualitatively, sound characteristics were preserved following changes in pump speed. Recordings using dedicated equipment showed that HeartMate II sound comprises low-frequency components corresponding to pump impeller rotation, as well as high-frequency components due to a pulse width modulation of the electric power to the pump. These different signal components interact and result in a complicated frequency spectrum. The iPhone and iPod recordings could not reproduce the sounds as well as the dedicated equipment. In particular, lower frequencies were affected by outside disturbances. Discussion: This article outlines a systematic approach to LVAD sound analysis using signal processing methods to quantify and potentially detect changes, and describes some of the challenges, for example, with the use of inexpensive recording devices.

This note considers the identification of large-scale one-dimensional networks consisting of identical LTI dynamical systems. A subspace identification method is developed that only uses local input-output information and does not rely on knowledge about the local state interaction. The proposed identification method estimates the Markov parameters of a locally lifted system, following the state-space realization of a single subsystem. The Markov-parameter estimation is formulated as a rank minimization problem by exploiting the low-rank property and the two-layer Toeplitz structural property in the data equation, whereas the state-space realization of a single subsystem is formulated as a structured low-rank matrix-factorization problem. The effectiveness of the proposed identification method is demonstrated by simulation examples.

Time-correlated single-photon counting lidar provides very high-resolution range measurements, making the technology interesting for 3D imaging of objects behind foliage or other obscuration. We study six peak detection approaches and compare their performance from several perspectives: detection of double surfaces within the instantaneous field of view, range accuracy, performance under sparse sampling, and the number of outliers. The results presented are based on reference measurements of a characterization target. Special consideration is given to the possibility of resolving two surfaces closely separated in range within the field of view of a single pixel. An approach based on fitting a linear combination of impulse response functions to the collected data showed the best overall performance. (c) 2018 Society of Photo-Optical Instrumentation Engineers (SPIE).

This paper presents a new method to compute upper and lower bounds of any voltage or current of an arbitrary linear electric circuit model with uncertain parameters. The bounds are in the frequency domain, and when compared to a previously proposed method, this novel approach provides a higher level of guarantee. The reason is that the bounds are not only computed for a set of fixed frequencies but also computed to a set of intervals of frequencies. The details of the proposed approach, especially the equivalent uncertain element models, are given. Additionally, tests are performed on problems with low and high number of uncertain parameters. Contrary to the classical method of Monte Carlo, the results are not based on a random choice of parameters and do not depend on the number of iterations. It is shown on an example that the classical method of Monte Carlo needs a high number of iterations to reach results in agreement with the proposed method. Then, it leads to higher computation times of several orders of magnitude.

Nonlinear cooperative systems associated to vector fields that are concave or subhomogeneous describe well interconnected dynamics that are of key interest for communication, biological, economical, and neural network applications. For this class of positive systems, we provide conditions that guarantee existence, uniqueness and stability of strictly positive equilibria. These conditions can be formulated directly in terms of the spectral radius of the Jacobian of the system. If control inputs are available, then it is shown how to use state feedback to stabilize an equilibrium point in the interior of the positive orthant.

This study proposes a new model which is fully specified for automated seizure onset detection and seizure onset prediction based on electroencephalography (EEG) measurements. We processed two archetypal EEG databases, Freiburg (intracranial EEG) and CHB-MIT (scalp EEG), to find if our model could outperform the state-of-the art models. Four key components define our model: (1) multiscale principal component analysis for EEG de-noising, (2) EEG signal decomposition using either empirical mode decomposition, discrete wavelet transform or wavelet packet decomposition, (3) statistical measures to extract relevant features, (4) machine learning algorithms. Our model achieved overall accuracy of 100% in ictal vs. inter-ictal EEG for both databases. In seizure onset prediction, it could discriminate between inter-ictal, pre-ictal, and ictal EEG with the accuracy of 99.77%, and between inter-ictal and pre-ictal EEG states with the accuracy of 99.70%. The proposed model is general and should prove applicable to other classification tasks including detection and prediction regarding bio-signals such as EMG and ECG. (C) 2017 Elsevier Ltd. All rights reserved.

Merge scenarios confront drivers with some of the most complicated driving maneuvers in every day driving, requiring anticipatory reasoning of positions of other vehicles, and the own vehicles future trajectory. In congested traffic it might be impossible to merge without cooperation of up-stream vehicles, therefore, it is essential to gauge the effect of our own trajectory when planning a merge maneuver. For an autonomous vehicle to perform a merge maneuver in congested traffic similar capabilities are required. This includes a model describing the future evolution of the scene that allows for optimizing the autonomous vehicle's planned trajectory with respect to risk, comfort, and dynamical limitations. We present a probabilistic model that explicitly models interaction between vehicles and allows for evaluating the utility of a large number of candidate trajectories of an autonomous vehicle using a receding horizon approach in order to select an appropriate merge maneuver. The model is an extension of the intelligent driver model and the modeled behavior of other vehicles are adjusted using on-line model parameter estimation in order to give better predictions. The prediction model is evaluated using naturalistic traffic data and the merge maneuver planner is evaluated in simulation.

State-space smoothing has found many applications in science and engineering. Under linear and Gaussian assumptions, smoothed estimates can be obtained using efficient recursions, for example Rauch Tung Striebel and Mayne Fraser algorithms. Such schemes are equivalent to linear algebraic techniques that minimize a convex quadratic objective function with structure induced by the dynamic model. These classical formulations fall short in many important circumstances. For instance, smoothers obtained using quadratic penalties can fail when outliers are present in the data, and cannot track impulsive inputs and abrupt state changes. Motivated by these shortcomings, generalized Kalman smoothing formulations have been proposed in the last few years, replacing quadratic models with more suitable, often nonsmooth, convex functions. In contrast to classical models, these general estimators require use of iterated algorithms, and these have received increased attention from control, signal processing, machine learning, and optimization communities. In this survey we show that the optimization viewpoint provides the control and signal processing community great freedom in the development of novel modeling and inference frameworks for dynamical systems. We discuss general statistical models for dynamic systems, making full use of nonsmooth convex penalties and constraints, and providing links to important models in signal processing and machine learning. We also survey optimization techniques for these formulations, paying close attention to dynamic problem structure. Modeling concepts and algorithms are illustrated with numerical examples. (C) 2017 Elsevier Ltd. All rights reserved.

Predicting the sign of press perturbation responses in ecological networks is challenging, due to the poor knowledge of the strength of the direct interactions among the species, and to the entangled coexistence of direct and indirect effects. We show in this paper that, for a class of networks that includes mutualistic and monotone networks, the sign of press perturbation responses can be qualitatively determined based only on the sign pattern of the community matrix, without any knowledge of parameter values. For other classes of networks, we show that a semi-qualitative approach yields sufficient conditions for community matrices with a given sign pattern to exhibit mutualistic responses to press perturbations; quantitative conditions can be provided as well for community matrices that are eventually nonnegative. We also present a computational test that can be applied to any class of networks so as to check whether the sign of the responses to press perturbations is constant in spite of parameter variations.

n/a

In this paper, we study the problem of controlling complex networks with unilateral controls, i.e., controls which can assume only positive or negative values, not both. Given a complex network represented by the adjacency matrix A, an algorithm is developed that constructs an input matrix B such that the resulting system (A, B) is controllable with a near minimal number of unilateral control inputs. This is made possible by a reformulation of classical conditions for controllability that casts the minimal unilateral input selection problem into well known optimization problems. We identify network properties that make unilateral controllability relatively easy to achieve as compared to unrestricted controllability. The analysis of the network topology for instance allows us to establish theoretical bounds on the minimal number of controls required. For various categories of random networks as well as for a number of real-world networks these lower bounds are often achieved by our heuristics.

We propose a novel class of Sequential Monte Carlo (SMC) algorithms, appropriate for inference in probabilistic graphical models. This class of algorithms adopts a divide-and-conquer approach based upon an auxiliary tree-structured decomposition of the model of interest, turning the overall inferential task into a collection of recursively solved subproblems. The proposed method is applicable to a broad class of probabilistic graphical models, including models with loops. Unlike a standard SMC sampler, the proposed divide-and-conquer SMC employs multiple independent populations of weighted particles, which are resampled, merged, and propagated as the method progresses. We illustrate empirically that this approach can outperform standard methods in terms of the accuracy of the posterior expectation and marginal likelihood approximations. Divide-and-conquer SMC also opens up novel parallel implementation options and the possibility of concentrating the computational effort on the most challenging subproblems. We demonstrate its performance on a Markov random field and on a hierarchical logistic regression problem. Supplementary materials including proofs and additional numerical results are available online.

A platform for sensor fusion consisting of a standard smartphone equipped with the specially developed Sensor Fusion appis presented. The platform enables real-time streaming of data over WiFi to a computer where signal processingalgorithms, e.g., the Kalman filter, can be developed and executed in a Matlab framework. The platform is an excellenttool for educational purposes and enables learning activities where methods based on advanced theory can be implementedand evaluated at low cost. The article describes the app and a laboratory exercise developed around these new technologicalpossibilities. The laboratory session is part of a course in sensor fusion, a signal processing continuation course focused onmultiple sensor signal applications, where the goal is to give the students hands on experience of the subject. This is done byestimating the orientation of the smartphone, which can be easily visualized and also compared to the built-in filters in thesmartphone. The filter can accept any combination of sensor data from accelerometers, gyroscopes, and magnetometers toexemplify their importance. This way different tunings and tricks of important methods are easily demonstrated andevaluated on-line. The presented framework facilitates this in a way previously impossible.

Bayesian nonparametric approaches have been recently introduced in system identification scenario where the impulse response is modeled as the realization of a zero-mean Gaussian process whose covariance (kernel) has to be estimated from data. In this scheme, quality of the estimates crucially depends on the parametrization of the covariance of the Gaussian process. A family of kernels that have been shown to be particularly effective in the system identification framework is the family of Diagonal/Correlated (DC) kernels. Maximum entropy properties of a related family of kernels, the Tuned/Correlated (TC) kernels, have been recently pointed out in the literature. In this technical note, we show that maximum entropy properties indeed extend to the whole family of DC kernels. The maximum entropy interpretation can be exploited in conjunction with results on matrix completion problems in the graphical models literature to shed light on the structure of the DC kernel. In particular, we prove that the DC kernel admits a closed-form factorization, inverse, and determinant. These results can be exploited both to improve the numerical stability and to reduce the computational complexity associated with the computation of the DC estimator.

In this paper the main goal is to compare the instrumental variables and the least squares methods applied to parameter estimation in continuous-time systems, avoiding any preliminary discretization of the process, and to analyse which method is more suitable for estimation in continuous-time under stochastic perturbations. A numerical example illustrates the effectiveness of the algorithms.

n/a

The general Simultaneous Localisation and Mapping (SLAM) problem aims at estimating the state of a moving platform simultaneously with building a map of the local environment. There are essentially three classes of algorithms. EKF- SLAM and FastSLAM solve the problem on-line, while Nonlinear Least Squares (NLS) is a batch method. All of them scales badly with either the state dimension, the map dimension or the batch length. We investigate the EM algorithm for solving a generalized version of the NLS problem. This EM-SLAM algorithm solves two simpler problems iteratively, hence it scales much better with dimensions. The iterations switch between state estimation, where we propose an Extended Rauch-Tung-Striebel smoother, and map estimation, where a quasi-Newton method is suggested. The proposed method is evaluated in real experiments and also in simulations on a platform with a monocular camera attached to an inertial measurement unit. It is demonstrated to produce lower RMSE than with a standard Levenberg-Marquardt solver of NLS problem, at a computational cost that increases considerably slower. 

This paper reviews the vast literature on static output feedback design for linear time-invariant systems including classical results and recent developments. In particular, we focus on static output feedback synthesis with performance specifications, structured static output feedback, and robustness. The paper provides a comprehensive review on existing design approaches including iterative linear matrix inequalities heuristics, linear matrix inequalities with rank constraints, methods with decoupled Lyapunov matrices, and non-Lyapunov-based approaches. We describe the main difficulties of dealing with static output feedback design and summarize the main features, advantages, and limitations of existing design methods. (C) 2016 International Federation of Automatic Control. Published by Elsevier Ltd. All rights reserved.

Todays 4G LTE systems bring unprecedented mobile broadband performance to over a billion of users across the globe. Recently, work on a 5G mobile communication system has begun, and next to a new 5G air interface, LTE will be an essential component. The evolution of LTE will therefore strive to meet 5G requirements and to address 5G use cases. In this article, we provide an overview of foreseen key technology areas and components for LTE Release 14, including latency reductions, enhancements for machine-type communication, operation in unlicensed spectrum, massive multi-antenna systems, broadcasting, positioning, and support for intelligent transportation systems.

The identification of multivariable state space models in innovation form is solved in a subspace identification framework using convex nuclear norm optimization. The convex optimization approach allows to include constraints on the unknown matrices in the data-equation characterizing subspace identification methods, such as the lower triangular block-Toeplitz of weighting matrices constructed from the Markov parameters of the unknown observer. The classical use of instrumental variables to remove the influence of the innovation term on the data equation in subspace identification is avoided. The avoidance of the instrumental variable projection step has the potential to improve the accuracy of the estimated model predictions, especially for short data length sequences. (C) 2016 Elsevier Ltd. All rights reserved.

This paper presents a data-driven receding horizon fault estimation method for additive actuator and sensor faults in unknown linear time-invariant systems, with enhanced robustness to stochastic identification errors. State-of-the-art methods construct fault estimators with identified state-space models or Markov parameters, without compensating for identification errors. Motivated by this limitation, we first propose a receding horizon fault estimator parameterized by predictor Markov parameters. This estimator provides (asymptotically) unbiased fault estimates as long as the subsystem from faults to outputs has no unstable transmission zeros. When the identified Markov parameters are used to construct the above fault estimator, stochastic identification errors appear as model uncertainty multiplied with unknown fault signals and online system inputs/outputs (I/O). Based on this fault estimation error analysis, we formulate a mixed-norm problem for the offline robust design that regards online I/O data as unknown. An alternative online mixed-norm problem is also proposed that can further reduce estimation errors at the cost of increased computational burden. Based on a geometrical interpretation of the two proposed mixed-norm problems, systematic methods to tune the user-defined parameters therein are given to achieve desired performance trade-offs. Simulation examples illustrate the benefits of our proposed methods compared to recent literature. (C) 2016 Elsevier Ltd. All rights reserved.

Nonlinear Kalman filters are algorithms that approximately solve the Bayesian filtering problem by employing the measurement update of the linear Kalman filter (KF). Numerous variants have been developed over the past decades, perhaps most importantly the popular sampling based sigma point Kalman filters.In order to make the vast literature accessible, we present nonlinear KF variants in a common framework that highlights the computation of mean values and covariance matrices as the main challenge. The way in which these moment integrals are approximated distinguishes, for example, the unscented KF from the divided difference KF.With the KF framework in mind, a moment computation problem is defined and analyzed. It is shown how structural properties can be exploited to simplify its solution. Established moment computation methods, and their basics and extensions, are discussed in an extensive survey. The focus is on the sampling based rules that are used in sigma point KF. More specifically, we present three categories of methods that use sigma-points 1) to represent a distribution (as in the UKF); 2) for numerical integration (as in Gauss-Hermite quadrature); 3) to approximate nonlinear functions (as in interpolation). Prospective benefits and downsides are listed for each of the categories and methods, including accuracy statements. Furthermore, the related KF publications are listed.The theoretical discussion is complemented with a comparative simulation study on instructive examples.

In this study, Random Forests (RF) classifier is proposed for ECG heartbeat signal classification in diagnosis of heart arrhythmia. Discrete wavelet transform (DWT) is used to decompose ECG signals into different successive frequency bands. A set of different statistical features were extracted from the obtained frequency bands to denote the distribution of wavelet coefficients. This study shows that RF classifier achieves superior performances compared to other decision tree methods using 10-fold cross-validation for the ECG datasets and the obtained results suggest that further significant improvements in terms of classification accuracy can be accomplished by the proposed classification system. Accurate ECG signal classification is the major requirement for detection of all arrhythmia types. Performances of the proposed system have been evaluated on two different databases, namely MIT-BIH database and St. -Petersburg Institute of Cardiological Technics 12-lead Arrhythmia Database. For MIT-BIH database, RF classifier yielded an overall accuracy 99.33 % against 98.44 and 98.67 % for the C4.5 and CART classifiers, respectively. For St. -Petersburg Institute of Cardiological Technics 12-lead Arrhythmia Database, RF classifier yielded an overall accuracy 99.95 % against 99.80 % for both C4.5 and CART classifiers, respectively. The combined model with multiscale principal component analysis (MSPCA) de-noising, discrete wavelet transform (DWT) and RF classifier also achieves better performance with the area under the receiver operating characteristic (ROC) curve (AUC) and F- measure equal to 0.999 and 0.993 for MIT-BIH database and 1 and 0.999 for and St. Petersburg Institute of Cardiological Technics 12-lead Arrhythmia Database, respectively. Obtained results demonstrate that the proposed system has capacity for reliable classification of ECG signals, and to assist the clinicians for making an accurate diagnosis of cardiovascular disorders (CVDs).

The first order stable spline (SS-1) kernel (also known as the tunedcorrelated kernel) is used extensively in regularized system identification, where the impulse response is modeled as a zero-mean Gaussian process whose covariance function is given by well designed and tuned kernels. In this paper, we discuss the maximum entropy properties of this kernel. In particular, we formulate the exact maximum entropy problem solved by the SS-1 kernel without Gaussian and uniform sampling assumptions. Under general sampling assumption, we also derive the special structure of the SS-1 kernel (e.g. its tridiagonal inverse and factorization have closed form expression), also giving to it a maximum entropy covariance completion interpretation.

This note presents an efficient approach for the evaluation of multi-parametric mixed integer quadratic programming (mp-MIQP) solutions, occurring for instance in control problems involving discrete time hybrid systems with quadratic cost. Traditionally, the online evaluation requires a sequential comparison of piecewise quadratic value functions. We introduce a lifted parameter space in which the piecewise quadratic value functions become piecewise affine and can be merged to a single value function defined over a single polyhedral partition without any overlaps. This enables efficient point location approaches using a single binary search tree. Numerical experiments with a power electronics application demonstrate an online speedup up to an order of magnitude. We also show how the achievable online evaluation time can be traded off against the offline computational time.

Safety and security applications benefit from better situational awareness. Radar micro-Doppler signatures from an observed target carry information about the target's activity, and have potential to improve situational awareness. This article describes, compares, and discusses two methods to classify human activity based on radar micro-Doppler data. The first method extracts physically interpretable features from the time-velocity domain such as the main cycle time and properties of the envelope of the micro-Doppler spectra and use these in the classification. The second method derives its features based on the components with the most energy in the cadence-velocity domain (obtained as the Fourier transform of the time-velocity domain). Measurements from a field trial show that the two methods have similar activity classification performance. It is suggested that target base velocity and main limb cadence frequency are indirect features of both methods, and that they do often alone suffice to discriminate between the studied activities. This is corroborated by experiments with a reduced feature set. This opens up for designing new more compact feature sets. Moreover, weaknesses of the methods and the impact of non-radial motion are discussed.

Tracking human body motions using inertial sensors has become a well-accepted method in ambulatory applications since the subject is not confined to a lab-bounded volume. However, a major drawback is the inability to estimate relative body positions over time because inertial sensor information only allows position tracking through strapdown integration, but does not provide any information about relative positions. In addition, strapdown integration inherently results in drift of the estimated position over time. We propose a novel method in which a permanent magnet combined with 3-D magnetometers and 3-D inertial sensors are used to estimate the global trunk orientation and relative pose of the hand with respect to the trunk. An Extended Kalman Filter is presented to fuse estimates obtained from inertial sensors with magnetic updates such that the position and orientation between the human hand and trunk as well as the global trunk orientation can be estimated robustly. This has been demonstrated in multiple experiments in which various hand tasks were performed. The most complex task in which simultaneous movements of both trunk and hand were performed resulted in an average rms position difference with an optical reference system of 19.7 +/- 2.2 mm whereas the relative trunk-hand and global trunk orientation error was 2.3 +/- 0.9 and 8.6 +/- 8.7 deg respectively.

The particle Gibbs sampler is a systematic way of using a particle filter within Markov chain Monte Carlo. This results in an off-the-shelf Markov kernel on the space of state trajectories, which can be used to simulate from the full joint smoothing distribution for a state space model in a Markov chain Monte Carlo scheme. We show that the particle Gibbs Markov kernel is uniformly ergodic under rather general assumptions, which we will carefully review and discuss. In particular, we provide an explicit rate of convergence, which reveals that (i) for fixed number of data points, the convergence rate can be made arbitrarily good by increasing the number of particles and (ii) under general mixing assumptions, the convergence rate can be kept constant by increasing the number of particles superlinearly with the number of observations. We illustrate the applicability of our result by studying in detail a common stochastic volatility model with a non-compact state space.

In this paper, we propose using Gaussian processes to track an extended object or group of objects, that generates multiple measurements at each scan. The shape and the kinematics of the object are simultaneously estimated, and the shape is learned online via a Gaussian process. The proposed algorithm is capable of tracking different objects with different shapes within the same surveillance region. The shape of the object is expressed analytically, with well-defined confidence intervals, which can be used for gating and association. Furthermore, we use an efficient recursive implementation of the algorithm by deriving a state space model in which the Gaussian process regression problem is cast into a state estimation problem.

Filtering and smoothing algorithms for linear discrete-time state-space models with skewed and heavy-tailed measurement noise are presented. The algorithms use a variational Bayes approximation of the posterior distribution of models that have normal prior and skew-t-distributed measurement noise. The proposed filter and smoother are compared with conventional low-complexity alternatives in a simulated pseudorange positioning scenario. In the simulations the proposed methods achieve better accuracy than the alternative methods, the computational complexity of the filter being roughly 5 to 10 times that of the Kalman filter.

Today, the workflows that are involved in industrial assembly and production activities are becoming increasingly complex. To efficiently and safely perform these workflows is demanding on the workers, in particular when it comes to infrequent or repetitive tasks. This burden on the workers can be eased by introducing smart assistance systems. This article presents a scalable concept and an integrated system demonstrator designed for this purpose. The basic idea is to learn workflows from observing multiple expert operators and then transfer the learnt workflow models to novice users. Being entirely learning-based, the proposed system can be applied to various tasks and domains. The above idea has been realized in a prototype, which combines components pushing the state of the art of hardware and software designed with interoperability in mind. The emphasis of this article is on the algorithms developed for the prototype: 1) fusion of inertial and visual sensor information from an on-body sensor network (BSN) to robustly track the user’s pose in magnetically polluted environments; 2) learning-based computer vision algorithms to map the workspace, localize the sensor with respect to the workspace and capture objects, even as they are carried; 3) domain-independent and robust workflow recovery and monitoring algorithms based on spatiotemporal pairwise relations deduced from object and user movement with respect to the scene; and 4) context-sensitive augmented reality (AR) user feedback using a head-mounted display (HMD). A distinguishing key feature of the developed algorithms is that they all operate solely on data from the on-body sensor network and that no external instrumentation is needed. The feasibility of the chosen approach for the complete action-perception-feedback loop is demonstrated on three increasingly complex datasets representing manual industrial tasks. These limited size datasets indicate and highlight the potential of the chosen technology as a combined entity as well as point out limitations of the system.

We study cooperative sensor network localization in a realistic scenario where 1) the underlying measurement errors more probably follow a non-Gaussian distribution; 2) the measurement error distribution is unknown without conducting massive offline calibrations; and 3) non-line-of-sight identification is not performed due to the complexity constraint and/or storage limitation. The underlying measurement error distribution is approximated parametrically by a Gaussian mixture with finite number of components, and the expectation-conditional maximization (ECM) criterion is adopted to approximate the maximum-likelihood estimator of the unknown sensor positions and an extra set of Gaussian mixture model parameters. The resulting centralized ECM algorithms lead to easier inference tasks and meanwhile retain several convergence properties with a proof of the "space filling" condition. To meet the scalability requirement, we further develop two distributed ECM algorithms where an average consensus algorithm plays an important role for updating the Gaussian mixture model parameters locally. The proposed algorithms are analyzed systematically in terms of computational complexity and communication overhead. Various computer based tests are also conducted with both simulation and experimental data. The results pin down that the proposed distributed algorithms can provide overall good performance for the assumed scenario even under model mismatch, while the existing competing algorithms either cannot work without the prior knowledge of the measurement error statistics or merely provide degraded localization performance when the measurement error is clearly non-Gaussian.

For communities of agents which are not necessarily cooperating, distributed processes of opinion forming are naturally represented by signed graphs, with positive edges representing friendly and cooperative interactions and negative edges the corresponding antagonistic counterpart. Unlike for nonnegative graphs, the outcome of a dynamical system evolving on a signed graph is not obvious and it is in general difficult to characterize, even when the dynamics are linear. In this paper, we identify a significant class of signed graphs for which the linear dynamics are however predictable and show many analogies with positive dynamical systems. These cases correspond to adjacency matrices that are eventually positive, for which the Perron-Frobenius property still holds and implies the existence of an invariant cone contained inside the positive orthant. As examples of applications, we determine cases in which it is possible to anticipate or impose unanimity of opinion in decision/voting processes even in presence of stubborn agents, and show how it is possible to extend the PageRank algorithm to include negative links.

In optimization algorithms used for on-line Model Predictive Control (MPC), linear systems of equations are often solved in each iteration. This is true both for Active Set methods as well as for Interior Point methods, and for linear MPC as well as for nonlinear MPC and hybrid MPC. The main computational effort is spent while solving these linear systems of equations, and hence, it is of greatest interest to solve them efficiently. Classically, the optimization problem has been formulated in either of two ways. One leading to a sparse linear system of equations involving relatively many variables to compute in each iteration and another one leading to a dense linear system of equations involving relatively few variables. In this work, it is shown that it is possible not only to consider these two distinct choices of formulations. Instead it is shown that it is possible to create an entire family of formulations with different levels of sparsity and number of variables, and that this extra degree of freedom can be exploited to obtain even better performance with the software and hardware at hand. This result also provides a better answer to a recurring question in MPC; should the sparse or dense formulation be used.

This paper addresses one of the main challenges in physical activity monitoring, as indicated by recent benchmark results: The difficulty of the complex classification problems exceeds the potential of existing classifiers. Therefore, this paper proposes the ConfAdaBoost.M1 algorithm. This algorithm is a variant of the AdaBoost.M1 that incorporates well-established ideas for confidence-based boosting. ConfAdaBoost.M1 is compared to the most commonly used boosting methods using benchmark datasets from the UCI machine learning repository.  Moreover, it is evaluated on an activity recognition and an intensity estimation problem, including a large number of physical activities from the recently released PAMAP2 dataset. The presented results indicate that the proposed ConfAdaBoost.M1 algorithm significantly improves the classification performance on most of the evaluated datasets, especially for larger and more complex classification tasks. Finally, two empirical studies are designed and carried out to investigate the feasibility of ConfAdaBoost.M1 for physical activity monitoring applications in mobile systems.

Particle Metropolis-Hastings (PMH) allows for Bayesian parameter inference in nonlinear state space models by combining MCMC and particle filtering. The latter is used to estimate the intractable likelihood. In its original formulation, PMH makes use of a marginal MCMC proposal for the parameters, typically a Gaussian random walk. However, this can lead to a poor exploration of the parameter space and an inefficient use of the generated particles.

We propose two alternative versions of PMH that incorporate gradient and Hessian information about the posterior into the proposal. This information is more or less obtained as a byproduct of the likelihood estimation. Indeed, we show how to estimate the required information using a fixed-lag particle smoother, with a computational cost growing linearly in the number of particles. We conclude that the proposed methods can: (i) decrease the length of the burn-in phase, (ii) increase the mixing of the Markov chain at the stationary phase, and (iii) make the proposal distribution scale invariant which simplifies tuning.

A method for fusing Synthetic Aperture Radar (SAR) images with opticalaerial images is presented. This is done in a navigation framework, where the absolute position and orientation of the flying platform, as computed from the inertial navigation system, is corrected based on the aerial image coordinates taken as ground truth. The method is suitable for new low-price SAR systems for small unmanned vehicles. The primary application is remote sensing, where the SAR image provides one further "colour" channel revealing reflectivity to radio waves. The method is based on first applying an edge detection algorithm to the images and then optimising the most important navigation states by matching the two binary images. To get a measure of the estimation uncertainty, we embed the optimisation in a least squares framework, where an explicit method to estimate the (relative) size of the errors is presented. The performance is demonstrated on real SAR and aerial images, leading to an error of only a few pixels.

Estimation-based iterative learning control (ILC) is applied to a parallel kinematic manipulator known as the Gantry-Tau parallel robot. The system represents a control problem where measurements of the controlled variables are not available. The main idea is to use estimates of the controlled variables in the ILC algorithm, and in the paper this approach is evaluated experimentally on the Gantry-Tau robot. The experimental results show that an ILC algorithm using estimates of the tool position gives a considerable improvement of the control performance. The tool position estimate is obtained by fusing measurements of the actuator angular positions with measurements of the tool path acceleration using a complementary filter.

Particle Markov chain Monte Carlo (PMCMC) is a systematic way of combining the two main tools used for Monte Carlo statistical inference: sequential Monte Carlo (SMC) and Markov chain Monte Carlo (MCMC). We present a new PMCMC algorithm that we refer to as particle Gibbs with ancestor sampling (PGAS). PGAS provides the data analyst with an off-the-shelf class of Markov kernels that can be used to simulate, for instance, the typically high-dimensional and highly autocorrelated state trajectory in a state-space model. The ancestor sampling procedure enables fast mixing of the PGAS kernel even when using seemingly few particles in the underlying SMC sampler. This is important as it can significantly reduce the computational burden that is typically associated with using SMC. PGAS is conceptually similar to the existing PG with backward simulation (PGBS) procedure. Instead of using separate forward and backward sweeps as in PGBS, however, we achieve the same effect in a single forward sweep. This makes PGAS well suited for addressing inference problems not only in state-space models, but also in models with more complex dependencies, such as non-Markovian, Bayesian nonparametric, and general probabilistic graphical models.

Among the many different formulations of Model Predictive Control (MPC) with guaranteed stability, one that has attracted significant attention is the formulation with a terminal cost and terminal constraint set, the so called dual mode formulation. In this technical note our goal is to make minimal changes to the dual mode framework, for the linear polytopic case, in order to develop a flexible reference tracking algorithm with guaranteed stability and low complexity, which is intuitive and easily understood. The main idea is to introduce a scaling variable that dynamically scales the terminal constraint set and therefore allows it to be centered around an arbitrary setpoint without violating the stability conditions. The main benefit of the algorithm is reduced complexity of the resulting QP compared to other state of art methods without loosing performance.

The electroencephalogram (EEG) signal is very important in the diagnosis of epilepsy. Long-term EEG recordings of an epileptic patient contain a huge amount of EEG data. The detection of epileptic activity is, therefore, a very demanding process that requires a detailed analysis of the entire length of the EEG data, usually performed by an expert. This paper describes an automated classification of EEG signals for the detection of epileptic seizures using wavelet transform and statistical pattern recognition. The decision making process is comprised of three main stages: (a) feature extraction based on wavelet transform, (b) feature space dimension reduction using scatter matrices and (c) classification by quadratic classifiers. The proposed methodology was applied on EEG data sets that belong to three subject groups: (a) healthy subjects, (b) epileptic subjects during a seizure-free interval and (c) epileptic subjects during a seizure. An overall classification accuracy of 99% was achieved. The results confirmed that the proposed algorithm has a potential in the classification of EEG signals and detection of epileptic seizures, and could thus further improve the diagnosis of epilepsy.

This paper presents a data-driven approach to diagnostics of systems that operate in a repetitive manner. Considering that data batches collected from a repetitive operation will be similar unless in the presence of an abnormality, a condition change is inferred by comparing the monitored data against an available nominal batch. The method proposed considers the comparison of data in the distribution domain, which reveals information of the data amplitude. This is achieved with the use of kernel density estimates and the Kullback–Leibler distance. To decrease sensitivity to disturbances while increasing sensitivity to faults, the use of a weighting vector is suggested which is chosen based on a labeled dataset. The framework is simple to implement and can be used without process interruption, in a batch manner. The approach is demonstrated with successful experimental and simulation applications to wear diagnostics in an industrial robot gearbox and for diagnostics of gear faults in a rotating machine.

We propose a new method for generating semidefinite relaxations of optimal power flow problems. The method is based on chordal conversion techniques: by dropping some equality constraints in the conversion, we obtain semidefinite relaxations that are computationally cheaper, but potentially weaker, than the standard semidefinite relaxation. Our numerical results show that the new relaxations often produce the same results as the standard semidefinite relaxation, but at a lower computational cost.

A method to identify linear parameter varying models through minimisation of an -norm objective is presented. The method uses a direct nonlinear programming approach to a non-convex problem. The reason to use -norm is twofold. To begin with, it is a well-known and widely used system norm, and second, the cost functions described in this paper become differentiable when using the -norm. This enables us to have a measure of first-order optimality and to use standard quasi-Newton solvers to solve the problem. The specific structure of the problem is utilised in great detail to compute cost functions and gradients efficiently. Additionally, a regularised version of the method, which also has a nice computational structure, is presented. The regularised version is shown to have an interesting interpretation with connections to worst-case approaches.

In this paper, the measurements of individual wheel speeds and the absolute position from a global positioning system are used for high-precision estimation of vehicle tire radii. The radii deviation from its nominal value is modeled as a Gaussian random variable and included as noise components in a simple vehicle motion model. The novelty lies in a Bayesian approach to estimate online both the state vector and the parameters representing the process noise statistics using a marginalized particle filter (MPF). Field tests show that the absolute radius can be estimated with submillimeter accuracy. The approach is tested in accordance with regulation 64 of the United Nations Economic Commission for Europe on a large data set (22 tests, using two vehicles and 12 different tire sets), where tire deflations are successfully detected, with high robustness, i.e., no false alarms. The proposed MPF approach outperforms common Kalman-filter-based methods used for joint state and parameter estimation when compared with respect to accuracy and robustness.

A marginal version of the enumeration Bayesian Cramer-Rao Bound (EBCRB) for jump Markov systems is proposed. It is shown that the proposed bound is at least as tight as EBCRB and the improvement stems from better handling of the nonlinearities. The new bound is illustrated to yield tighter results than BCRB and EBCRB on a benchmark example.

In this article we present a parametric branch and bound algorithm for computation of optimal and suboptimal solutions to parametric mixed-integer quadratic programs and parametric mixed-integer linear programs. The algorithm returns an optimal or suboptimal parametric solution with the level of suboptimality requested by the user. An interesting application of the proposed parametric branch and bound procedure is suboptimal explicit MPC for hybrid systems, where the introduced user-defined suboptimality tolerance reduces the storage requirements and the online computational effort, or even enables the computation of a suboptimal MPC controller in cases where the computation of the optimal MPC controller would be intractable. Moreover, stability of the system in closed loop with the suboptimal controller can be guaranteed a priori.

The norm-optimal iterative learning control (ilc) algorithm for linear systems is extended to an estimation-based norm-optimal ilc  algorithm where the controlled variables are not directly available as measurements. A separation lemma is presented, stating that if a stationary Kalman filter is used for linear time-invariant systems then the ilc  design is independent of the dynamics in the Kalman filter. Furthermore, the objective function in the optimisation problem is modified to incorporate the full probability density function of the error. Utilising the Kullback–Leibler divergence leads to an automatic and intuitive way of tuning the ilc  algorithm. Finally, the concept is extended to non-linear state space models using linearisation techniques, where it is assumed that the full state vector is estimated and used in the ilc  algorithm. Stability and convergence properties for the proposed scheme are also derived.

The dependence of radio signal propagation on the environment is  well known, and both statistical and deterministic methods have been presented in the literature. Such methods are either based on randomised or actual reflectors of radio signals. In this work, we instead aim at estimating the location of the reflectors based on geo-localised radio channel impulse reponse measurements and using methods from synthetic aperture radar (SAR). Radio channel data measurements from 3GPP E-UTRAN have been used to verify the usefulness of the proposed approach. The obtained images show that  the estimated reflectors are well correlated with the aerial map of the environment. Also, which part of the trajectory contributed to different reflectors have been estimated with promising results.

With the electromagnetic theory as basis, we present a sensor model for three-axis magnetometers suitable for localization and tracking as required in intelligent transportation systems and security applications. The model depends on a physical magnetic dipole model of the target and its relative position to the sensor. Both point target and extended target models are provided as well as a target orientation dependent model. The suitability of magnetometers for tracking is analyzed in terms of local observability and the Cramér Rao lower bound as a function of the sensor positions in a two sensor scenario. The models are validated with real field test data taken from various road vehicles which indicate excellent localization as well as identification of the magnetic target model suitable for target classification. These sensor models can be combined with a standard motion model and a standard nonlinear filter to track metallic objects in a magnetometer network.

Monte Carlo methods, in particular those based on Markov chains and on interacting particle systems, are by now tools that are routinely used in machine learning. These methods have had a profound impact on statistical inference in a wide range of application areas where probabilistic models are used. Moreover, there are many algorithms in machine learning which are based on the idea of processing the data sequentially, first in the forward direction and then in the backward direction. In this tutorial we will review a branch of Monte Carlo methods based on the forward-backward idea, referred to as backward simulators. These methods are useful for learning and inference in probabilistic models containing latent stochastic processes. The theory and practice of backward simulation algorithms have undergone a significant development in recent years and the algorithms keep finding new applications. The foundation for these methods is sequential Monte Carlo (SMC). SMC-based backward simulators are capable of addressing smoothing problems in sequential latent variable models, such as general, nonlinear/non-Gaussian state-space models (SSMs). However, we will also clearly show that the underlying backward simulation idea is by no means restricted to SSMs. Furthermore, backward simulation plays an important role in recent developments of Markov chain Monte Carlo (MCMC) methods. Particle MCMC is a systematic way of using SMC within MCMC. In this framework, backward simulation gives us a way to significantly improve the performance of the samplers. We review and discuss several related backward-simulation-based methods for state inference as well as learning of static parameters, both using a frequentistic and a Bayesian approach.

Localization is an enabling technology in many applications and services today and in the future. Satellite navigation often works fine for navigation, infotainment and location based services, and it is today the dominating solution in commercial products. A nice exception is the localization in Google Maps, where radio signal strength from WiFi and cellular networks are used as complementary information to increase accuracy and integrity. With the on-going trend with more autonomous functions being introduced in our vehicles and with all our connected devices, most of them operated in indoor enviroments where satellite signals are not available,there is an acute need for new solutions.

At the same time, our smartphones are getting more sophisticated in their sensor configuration. Therefore, in this chapter we present a freely available Sensor Fusion app developed in-house, how it works, how it has been used, and how it can be used based on a variety of applications in our research and student projects.

It is a well-known fact that exercising helps people improve their overall well-being; both physiological and psychological health. Regular moderate physical activity improves the risk of disease progression, improves the chances for successful rehabilitation, and lowers the levels of stress hormones. Physical fitness can be categorized in cardiovascular fitness, and muscular strength and endurance. A proper balance between aerobic activities and strength exercises are important to maximize the positive effects. This balance is not always easily obtained, so assistance tools are important. Hence, ambient assisted living (AAL) systems that support and motivate balanced training are desirable. This chapter presents methods to provide this, focusing on the methodologies and concepts implemented by the authors in the physical activity monitoring for aging people (PAMAP) platform. The chapter sets the stage for an architecture to provide personalized activity monitoring using a network of wearable sensors, mainly inertial measurement units (IMU). The main focus is then to describe how to do this in a personalizable way: (1) monitoring to provide an estimate of aerobic activities performed, for which a boosting based method to determine activity type, intensity, frequency, and duration is given; (2) supervise and coach strength activities. Here, methodologies are described for obtaining the parameters needed to provide real-time useful feedback to the user about how to exercise safely using the right technique.

In this chapter parallel implementations of hybrid MPC will be discussed. Different methods for achieving parallelism at different levels of the algorithms will be surveyed. It will be seen that there are many possible ways of obtaining parallelism for hybrid MPC, and it is by no means clear which possibilities that should be utilized to achieve the best possible performance. To answer this question is a challenge for future research.

In this paper we show how to synthesize simple explicit MPC controllers based on approximate dynamic programming. Here, a given MPC optimization problem over a finite horizon is solved iteratively as a series of problems of size one. The optimal cost function of each subproblem is approximated by a quadratic function that serves as a cost-to-go function for the subsequent iteration. The approximation is designed in such a way that closed-loop stability and recursive feasibility is maintained. Specifically, we show how to employ sum-of-squares relaxations to enforce that the approximate cost-to-go function is bounded from below and from above for all points of its domain. By resorting to quadratic approximations, the complexity of the resulting explicit MPC controller is considerably reduced both in terms of memory as well as the on-line computations. The procedure is applied to control an inverted pendulum and experimental data are presented to demonstrate viability of such an approach.

In this work we consider a nonlinear interconnected system describing a decision-making process in a community of agents characterized by the coexistence of collaborative and antagonistic interactions. The resulting signed graph is in general not structurally balanced. It is shown in the paper that the decision-making process is affected by the frustration of the signed graph, in the sense that a nontrivial decision can be reached only if the social commitment of the agents is high enough to win the disorder introduced by the frustration in the network. The higher the frustration of the graph, the higher the commitment strength required from the agents.

In this paper, Bayesian lower bounds (BLBs) are obtained via a general form of the Pythagorean theorem where the inner product derives from the joint or the a-posteriori probability density function (pdf). When joint pdf is considered, the BLBs obtained encompass the Weiss-Weinstein family (WWF). When a-posteriori pdf is considered, by resorting to an embedding between two ad hoc subspaces, it is shown that any ”standard” BLBs of the WWF admits a ”tighter” form which upper bounds the ”standard” form. Interestingly enough, this latter result may explain why the ”standard” BLBs of the WWF are not always as tight as expected, as exemplified in the case of the Bayesian Cram´er-Rao Bound. As a consequence an updated definition of efficiency is proposed, as well as the introduction of an updated class of efficient estimators.

This paper presents a systematic approach for computing local solutions to motion planning problems in non-convex environments using numerical optimal control techniques. It extends the range of use of state-of-the-art numerical optimal control tools to problem classes where these tools have previously not been applicable. Today these problems are typically solved using motion planners based on randomized or graph search. The general principle is to define a homotopy that transforms, or preferably relaxes, the original problem to an easily solved problem. In this work, it is shown that by combining a Sequential Quadratic Programming (SQP) method with a homotopy approach that gradually transforms the problem from a relaxed one to the original one, practically relevant locally optimal solutions to the motion planning problem can be computed. The approach is demonstrated in motion planning problems in challenging 2D and 3D environments, where the presented method significantly outperforms both a state-of-the-art numerical optimal control method and a state-of-the-art open-source optimizing sampling-based planner commonly used as benchmark. 

An experimental comparison of two feed-forward based frictioncompensation methods is presented. The first method is based on theLuGre friction model, using identified friction model parameters, andthe second method is based on B-spline network, where the networkweights are learned from experiments. The methods are evaluated andcompared via experiments using a six axis industrial robot carryingout circular movements of different radii. The experiments show thatthe learning-based friction compensation gives an error reduction ofthe same magnitude as for the LuGre-based friction compensation.

The number of sensors used in tracking scenariosis constantly increasing, this puts high demands on the trackingmethods to handle these data streams. Central processing (ideallyoptimal) puts high demands on the central node, is sensitive toinaccurate sensor parameters, and suffers from the single pointof failure problem. Decentralizing the tracking can improve this,but may give considerable performance loss. The newly presentedinverse covariance intersection method, proven to be consistent,even under unknown track cross correlations, is benchmarkedagainst alternatives. Different track-to-track methods, includingsmoothed association over a window, are compared. A scenariowith objects tracked in multiple cameras, not necessarily opti-mized for tracking, are used to give realism to the evaluations.

In this paper, marginal versions of the Bayesian Bhattacharyya lower bound (BBLB), which is a tighter alternative to the classical Bayesian Cramer-Rao bound, for discrete-time filtering are proposed. Expressions for the second and third-order marginal BBLBs are obtained and it is shown how these can be approximately calculated using particle filtering. A simulation example shows that the proposed bounds predict the achievable performance of the filtering algorithms better.

We present two approaches for using imaging sensors on-board small unmanned aerial systems (UAS) for positioning and navigation. Two types of sensors are used; laser scanners and a camera operating in the visual wavelengths. The laser scanners produce sparse 3D data that are registered to produce a local map. For the images from the video camera the optical flow and height estimates are fused and then matched with a geo-referenced aerial image. Both approaches include data from the inertial navigation system. The approaches can be used for accurate ego-positioning, and thus for navigation. The approaches are GPS independent and can work in GPS denied conditions, for example urban canyons, indoor environments, forest areas or while jammed. Applications are primarily within societal security and military defense.

Regularization of simple linear regression models for system identification is a recent much-studied problem. Several parameterizations ("kernels") of the regularization matrix have been suggested together with different ways of estimating ("tuning") its parameters. This contribution defines an asymptotic view on the problem of tuning and selection of kernels. It is shown that the SURE approach to parameter tuning provides an asymptotically consistent estimate of the optimal (in a MSE sense) hyperparameters. At the same time it is shown that the common marginal likelihood (empirical Bayes) approach does not enjoy that property. (C) 2017, IFAC (International Federation of Automatic Control) Hosting by Elsevier Ltd. All rights reserved.

\This paper deals with the problem of controlling linear complex networks in an efficient way, i.e., with limited control energy. A general principle is provided, based on the eigenvalues of the network. It is shown numerically that the cost of controlling a network grows with the (absolute value of the) real part of the eigenvalues of the adjacency matrix. Constructive rules for driver node selection are also provided, based on the (weighted) topology of the network. In particular, we show that the key to have an energetically efficient driver node placement strategy is to use the skewness of the outdegree versus indegree distributions of the network, a topological property not associated before to controllability. (C) 2017, IFAC (International Federation of Automatic Control) Hosting by Elsevier Ltd. All rights reserved.

The aim of this paper is to investigate the problem of selecting driver nodes enabling the control of a network with minimal control energy. The networks we are interested in are coupled harmonic oscillators, i.e., networks in which the eigenvalues are all purely imaginary. For them, several criteria for driver node selection are presented, based on the different measures of control energy considered in this context. The constructive algorithms we develop for these criteria are normally solving the problem in a heuristic way, although in one case the exact solution can be computed efficiently regardless of size. (C) 2017, IFAC (International Federation of Automatic Control) Hosting by Elsevier Ltd. All rights reserved.

Input selection is an important and oftentimes difficult challenge in system identification. In order to achieve less complex models, irrelevant inputs should be methodically and correctly discarded before or under the estimation process. In this paper we introduce a novel method of input selection that is carried out as a natural extension in a subspace method. We show that the method robustly and accurately performs input selection at various noise levels and that it provides good model estimates. (C) 2017, IFAC (International Federation of Automatic Control) Hosting by Elsevier Ltd. All rights reserved.

This paper addresses reconstruction of linear dynamic networks from heterogeneous datasets. Those datasets consist of measurements from linear dynamical systems in multiple experiments subjected to different experimental conditions, e.g., changes/perturbations in parameters, disturbance or noise. A main assumption is that the Boolean structures of the underlying networks are the same in all experiments. The ARMAX model is adopted to parameterize the general linear dynamic network representation "Dynamical Structure Function" (DSF), which provides the Granger Causality graph as a special case. The network identification is performed by integrating all available datasets and promote group sparsity to assure both network sparsity and the consistency of Boolean structures over datasets. In terms of solving the problem, a treatment by the iterative reweighted l(1) method is used, together with its implementations via proximal methods and ADMM for large-dimensional networks. (C) 2017, IFAC (International Federation of Automatic Control) Hosting by Elsevier Ltd. All rights reserved.

In the consensus problem considered in this paper, each agent can impose a lower and an upper bound on the achievable consensus values. We show that if such state constraints are implemented by saturating the value transmitted to the neighboring nodes, the resulting constrained consensus problem must converge to the intersection of the intervals imposed by the individual agents.

This paper considers the input design of kernelbased regularization methods for LTI system identification. We first derive the Bayesian mean squared error matrix under the Bayesian perspective, and then use some typical scalar measures (e.g., the A-optimality, D-optimality, and E-optimality) as optimization criteria for the input design problem. Instead of directly solving the nonconvex optimization problem, we propose a two-step procedure. The first step is to solve a convex optimization and the second one is to determine the inverse image of a quadratic map. Both of these two steps can be solved efficiently by the proposed method and hence all the globally optimal inputs are found. In particular, we show that for some kernels, the optimal input under the D-optimality has an explicit expression.

The continuous-time subspace identification using state-variable filtering has been investigated for a long time. Due to the simple orthogonal basis functions that were adopted by the existing methods, the identification performance is quite sensitive to the selection of the system-dynamic parameter associated with an orthogonal basis. To cope with this problem, a subspace identification method using generalized orthonormal(Takenaka-Malmquist) basis functions is developed, which has the potential to perform better than the existing state-variable filtering methods since the adopted Takenaka-Malmquist basis has more degree of freedom in selecting the system-dynamic parameters. As a price for the flexibility of the generalized orthonormal bases, the transformed state-space model is time-varying or parameter-varying which cannot be identified using traditional subspace identification methods. To this end, a new subspace identification algorithm is developed by exploiting the structural properties of the time-variant system matrices, which is then validated by numerical simulations.

Network reconstruction has become particularly important in systems biology, and is now expected to deliver information on causality. Systems in nature are inherently nonlinear. However, for nonlinear dynamical systems with hidden states, how to give a useful definition of dynamic networks is still an open question. This paper presents a useful definition of Boolean dynamic networks for a large class of nonlinear systems. Moreover, a robust inference method is provided. The well-known Millar-10 model in systems biology is used as a numerical example, which provides the ground truth of causal networks for key mRNAs involved in eukaryotic circadian clocks. In addition, as second contribution of this paper, we suggest definitions of linear network identifiability, which helps to unify the available work on network identifiability.

One-dimensional Bayesian filtering and smoothing problems can be solved numerically using a number of algorithms, even in nonlinear and non-Gaussian cases. In this educational paper we advocate for the benefits of visualizing the obtained posterior densities as complement to, e.g., estimation error analysis. In addition to a review of Bayesian filtering and smoothing and the respective point mass and particle solutions, we devise a novel algorithm for filtering when the likelihood cannot be evaluated. Several instructive examples are discussed and easily adjustable matlab code is provided as complement to this paper.

In this paper, the derivation of the Bayesian Bhattacharyya bound for discrete-time filtering as proposed ina paper by Reece and Nicholson is revisited. It turns out that the results presented in the aforementioned contribution are incorrect, as some expectations appearing in the information matrix recursions are missing. This paper gives a generalized derivation of the N-th order Bayesian Bhattacharyya bound and presents corrected expressions for the case N = 2. A nonlinear toy example is used to illustrate the results

Classification of motion mode (walking, running, standing still) and device mode (hand-held, in pocket, in back-pack) is an enabler in personal navigation systems for the purpose of saving energy and design parameter settings and also for its own sake. Our main contribution is to publish one of the most extensive datasets for this problem, including inertial data from eight users, each one performing three pre-defined trajectories carrying four smartphones and seventeen inertial measurement units on the body. All kind of metadata is available such as the ground truth of all modes and position. A second contribution is the first study on a joint classifier of motion and device mode, respectively, where preliminary but promising results are presented.

Many positioning systems rely on accuratetime of arrival measurements. In this paper, we addressnot only the accuracy but also the relevance of Time ofArrival (TOA) measurement error modeling. We discusshow better knowledge of these errors can improve relativedistance estimation, and compare the impact of differentlydetailed measurement error information. These models arecompared in simulations based on models derived froman Ultra Wideband (UWB) measurement campaign. Theconclusion is that significant improvements can be madewithout providing detailed received signal information butwith a generic and relevant measurement error model.

Advances in sensor systems have resulted in the availability of high resolution sensors, capable of generating massive amounts of data. For complex systems to run online, the primary focus is on computationally efficient filters for the estimation of latent states related to the data. In this paper a novel method for efficient state estimation with the unscented Kalman Filter is proposed. The focus is on applications consisting of a massive amount of data. From a modelling perspective, this amounts to a measurement vector with dimensionality significantly greater than the dimensionality of the state vector. The efficiency of the filter is derived from a parallel filter structure which is enabled by the expectation propagation algorithm. A novel parallel measurement processing expectation propagation unscented Kalman filter is developed. The primary advantage of the novel algorithm is in the ability to achieve computational improvements with negligible loses in filter accuracy. An example of robot localization with a high resolution laser rangefinder sensor is presented. A 47.53% decrease in computational time was exhibited for a scenario with a processing platform consisting of 4 processors, with a negligible loss in accuracy.

In polar region operations, drift sea ice positioning and tracking is useful for both scientific and safety reasons. Modeling ice movements has proven difficult, not least due to the lack of information of currents and winds of high enough resolution. Thus, observations of drift ice is essential to an up-to-date ice-tracking estimate.

Recent years have seen the rise of Unmanned Aerial Systems (UAS) as a platform for geoobservation, and so too for the tracking of sea ice. Being a mobile platform, the research on UAS path-planning is extensive and usually involves an objective-function to minimize. For the purpose of observation however, the objective-function typically changes as observations are made along the path.

Further, the general problem involves multiple UAS and—in the case of sea ice tracking—vast geographical areas.

In this paper we discuss the architectural outline of a system capable of fusing data from multiple sources—UAS’s and others—as well as incorporating that data for both path-planning, sea ice movement prediction and target initialization. The system contains tracking of sea ice objects, situation map logic and is expandable as discussed with path-planning capabilities for closing the loop of optimizing paths for information acquisition.

Assessing the fundamental performance limitationsin Bayesian filtering can be carried out using the parametricCram´er-Rao bound (CRB). The parametric CRB puts a lowerbound on mean square error (MSE) matrix conditioned on aspecific state trajectory realization. In this work, we derive theparametric CRB for state-space models, where the measurementequation is modeled by a Gaussian process regression.These models appear, for instance in proximity report-basedpositioning, where proximity reports are obtained by hardthresholding of received signal strength (RSS) measurements, thatare modeled through Gaussian process regression. The proposedparametric CRB is evaluated on selected state trajectories andfurther compared with the positioning performance obtained bythe particle filter. The results corroborate that the positioningaccuracy achieved in this framework is close to the parametricCRB.

 In this work an Inertial Measurement Unit is used to improve tool position estimates for an ABB IRB 4600 industrial robot, starting from estimates based on motor angle forward kinematics. A Complementary Filter and an Extended Kalman Filter are investigated. The Complementary Filter is found to perform on par with the Extended Kalman Filter while having lower complexity both in the tuning process and the filtering computations.

Acoustic frequency tracking of a harmonic signalwith continuously varying frequency is considered. The Rao-Blackwellized point mass filter (RBPMF), previously proposed bythe authors for mechanical vibration tracking, is applied to the problem. The RBPMF is compared with two periodogram-based methods, and the similarities and differences between them are explained. Both experimental and simulation results in a Doppler frequency tracking scenario are presented, and the results show that the RBPMF can have significantly less estimation error than the competing methods.

In polar region operations, drift ice positioning and tracking is useful for both scientific and safety reasons. At its core is a Multi-Target Tracking (MTT) problem in which currents and winds make motion modeling difficult. One recent algorithm in the MTT field, employed in this paper, is the Labeled Multi-Bernoulli (LMB) filter. In particular, a proposed reformulation of the LMB equations exposes a structure which is exploited to propose a compact algorithm for the generation of the filter's posterior distribution. Further, spatial indexing is applied to the clustering process of the filter, allowing efficient separation of the filter into smaller, independent parts with lesser total complexity than that of an unclustered filter. Many types of sensors can be employed to generate detections of sea ice, and in this paper a recorded dataset from a Terrestrial Radar Interferometer (TRI) is used to demonstrate the application of the Spatially Indexed Labeled Multi-Bernoulli filter to estimate the currents of an observed area in Kongsfjorden, Svalbard.

Modern fighter aircraft require maximum control performance in order to have the upper hand in a dogfight or when they have to outmaneuver an enemy missile. Therefore pilots must be able to maneuver the aircraft very close to the limit of what it is capable of while at the same time focus on the tactical tasks of the mission. To enable this, modern flight control systems have automatic systems for angle of attack and load factor limiting.

One such design technique is reference or command governors. In this paper we investigate different ways of designing command governors for angle of attack and load factor limiting in in fighter aircraft flight control systems. We discuss different design approaches and their properties and implement one selected design in Saab's JAS 39 Gripen fighter aircraft simulation environment. 

The stable spline kernel and the diagonal correlated kernel are two kernels that have been tested extensively in kernel-based regularization methods for LTI system identification. As shown in our recent works, although these two kernels are introduced in different ways, they share some common features, e.g., they all belong to the class of exponentially convex locally stationary kernels, and state-space model induced kernels. In this work, we further show that similar to the derivation of the stable spline kernel, the continuous-time diagonal correlated kernel can be derived by applying the same "stable" coordinate change to a "generalized" first order spline kernel, and thus can be interpreted as a stable generalized first order spline kernel. This interpretation provides new facets to understand the properties of the diagonal correlated kernel. Due to this interpretation, new eigendecompositions, explicit expression of the norm, and new maximum entropy interpretation of the diagonal correlated kernel are derived accordingly.

Contractive interference functions are a subclass of the standard interference functions used in the design and analysis of distributed power control algorithms for wireless networks. Their peculiarity is that for the resulting positive system the existence and global asymptotic stability of a unique positive equilibrium point is guaranteed. In this paper we give an infinitesimal characterization of nonlinear contractiveinterference functions in terms of the spectral radius of the Jacobian linearization at any point in the positive orthant. The condition we obtain, that the spectral radius is always less than 1, extends to the nonlinear case an equivalent property of linear interference functions, and leads to a Jacobian characterization similar to the one commonly used in contraction analysis of nonlinear systems.

It is a well known fact that finite time optimal controllers, such as MPC do not necessarily result in closed loop stable systems. Within the MPC community it is common practice to add a final state constraint and/or a final state penalty in order to obtain guaranteed stability. However, for more advanced controller structures it can be difficult to show stability using these techniques. Additionally in some cases the final state constraint set consists of so many inequalities that the complexity of the MPC problem is too big for use in certain fast and time critical applications. In this paper we instead focus on deriving a tool for a-postiori analysis of the closed loop stability for linear systems controlled with MPC controllers. We formulate an optimisation problem that gives a sufficient condition for stability of the closed loop system and we show that the problem can be written as a Mixed Integer Linear Programming Problem (MILP).

Reversing with a dolly steered trailer configura- tion is a hard task for any driver without extensive training. In this work we present a motion planning and control framework that can be used to automatically plan and execute complicated manoeuvres. The unstable dynamics of the reversing general 2- trailer configuration with off-axle hitching is first stabilised by an LQ-controller and then a pure pursuit path tracker is used on a higher level giving a cascaded controller that can track piecewise linear reference paths. This controller together with a kinematic model of the trailer configuration is then used for forward simulations within a Closed-Loop Rapidly Exploring Random Tree framework to generate motion plans that are not only kinematically feasible but also include the limitations of the controller’s tracking performance when reversing. The approach is evaluated over a series of Monte Carlo simulations on three different scenarios and impressive success rates are achieved. Finally the approach is successfully tested on a small scale test platform where the motion plan is calculated and then sent to the platform for execution. 

The aim of this paper is to suggest a modification to the usual bounded confidence model of opinion dynamics, so that “changes of opinion” (intended as changes of the sign of the initial state) of an agent are never induced by the dynamics. In order to do so, a bipartite consensus model is used, endowing it with a confidence range. The resulting signed bounded confidence model has a state-dependent connectivity and a behavior similar to its standard counterpart, but in addition it preserves the sign of the opinions by “repelling away” opinions localized near the origin but on different sides with respect to 0.

In many practical applications of system identification, it is not feasible to measure both the inputs applied to the system as well as the output. In such situations, it is desirable to estimate both the inputs and the dynamics of the system simultaneously; this is known as the blind identification problem. In this paper, we provide a novel extension of subspace methods to the blind identification of multiple-input multiple-output linear systems. We assume that our inputs lie in a known subspace, and we are able to formulate the identification problem as rank constrained optimization, which admits a convex relaxation. We show the efficacy of this formulation with a numerical example.

In this paper a cascaded approach for stabilizationand path tracking of a general 2-trailer vehicle configurationwith an off-axle hitching is presented. A low level LinearQuadratic controller is used for stabilization of the internalangles while a pure pursuit path tracking controller is used ona higher level to handle the path tracking. Piecewise linearityis the only requirement on the control reference which makesthe design of reference paths very general. A Graphical UserInterface is designed to make it easy for a user to design controlreferences for complex manoeuvres given some representationof the surroundings. The approach is demonstrated with challengingpath following scenarios both in simulation and on asmall scale test platform.

In this paper, recent results on the evaluation of the Bayesian Cramer-Rao bound for jump Markov systems are presented. In particular, previous work is extended to jump Markov systems where the discrete mode variable enters into both the process and measurement equation, as well as where it enters exclusively into the measurement equation. Recursive approximations are derived with finite memory requirements as well as algorithms for checking the validity of these approximations are established. The tightness of the bound and the validity of its approximation is investigated on a couple of examples.

We still have very little knowledge about how ourbrains decouple different sound sources, which is known assolving the cocktail party problem. Several approaches; includingERP, time-frequency analysis and, more recently, regression andstimulus reconstruction approaches; have been suggested forsolving this problem. In this work, we study the problem ofcorrelating of EEG signals to different sets of sound sources withthe goal of identifying the single source to which the listener isattending. Here, we propose a method for finding the number ofparameters needed in a regression model to avoid overlearning,which is necessary for determining the attended sound sourcewith high confidence in order to solve the cocktail party problem.

In inertial motion capture, a multitude of body segments are equipped with inertial sensors, consisting of 3D accelerometers and 3D gyroscopes. Using an optimization-based approach to solve the motion capture problem allows for natural inclusion of biomechanical constraints and for modeling the connection of the body segments at the joint locations. The computational complexity of solving this problem grows both with the length of the data set and with the number of sensors and body segments considered. In this work, we present a scalable and distributed solution to this problem using tailored message passing, capable of exploiting the structure that is inherent in the problem. As a proof-of-concept we apply our algorithm to data from a lower body configuration. 

Radar and sonar provide information of both range and radial velocity to unknown objects. This is accomplished by emitting a signal waveform and computing the round trip time and Doppler shift. Estimation of the round trip time and Doppler shift is usually done separately without considering the couplings between these two object related quantities. The purpose of this contribution is to first model the amplitude, time shift and time scale of the returned signal in terms of the object related states range and velocity, and analyse the Cramér-Rao lower bound (CRLB) for the joint range and velocity estimation problem. One of the conclusions is that there is negative correlation between range and velocity. The maximum likelihood (ML) cost function also confirms this strong negative correlation. For target tracking applications, the use of the correct covariance matrix for the measurement vector gives a significant gain in information, compared to using the variance of range and velocity assuming independence. In other words, the knowledge of the correlation tells the tracking filter that a too large range measurement comes most likely with a too small velocity measurement, and vice versa. Experiments with sound pulses reflected in a wall indoors confirm the main conclusion of negative correlation.

GPS is widely used for localization and tracking, however traditional GPS receivers consume too much energy for many applications. This paper implements and evaluates the performance of a low-energy GPS prototype. The main difference is that a traditional GPS needs to sample signals transmitted by satellites for 30 seconds to estimate its position. Our prototype reduces this time by three orders of magnitude and it can compute positions from only 2 milliseconds of data. We present a new algorithm that increases robustness by filtering on estimated residuals instead of using an altitude database. In addition, we show that our new algorithm works with both fixed and moving targets. The solution consists of (1) a portable device that samples the GPS signal and (2) a server that utilizes Doppler navigation and Coarse Time Navigation to estimate positions. We performed tests in a wide variety of environments and situations. These tests show that our prototype provides a median positioning error of roughly 40 meters even when the GPS receiver is moving at 80 kilometres per hour.

Positioning in cellular networks is often based on mobile-assisted measurements of serving and neighboring base stations. Traditionally, positioning is considered to be enabled when the mobile provides measurements of three different base stations. In this paper, we additionally investigate positioning based on time series of Time Of Flight (TOF) and Time Difference of Arrival (TDOA) measurements gathered from two base stations with known positions, where the specific base stations involved depend on the trajectory of the mobile station.. The set of two base stations is different along the trajectory. Each report contains TOF for the serving base station, and one TDOA measurement for the most favorable neighboring base station relative the serving base station. We derive explicit analytical solution related to the intersection of the absolute distance circle (from TOF) and relative distance hyperbola (from TDOA). We consider both geometric noise-free problem and the more realistic problem with additive noise as delivered in the 3rd Generation Partnership Project (3GPP) Long-Term Evolution (LTE). Positioning performance is evaluated using the Cramer-Rao lower bound.

This paper proposes a diagonal covariance matrix approximation for Wide-Sense Stationary (WSS) signals with correlated Gaussian noise. Existing signal models that incorporate correlations often require regularization of the covariance matrix, so that the covariance matrix can be inverted. The disadvantage of this approach is that matrix inversion is computational intensive and regularization decreases precision. We use Bienayme's theorem to approximate the covariance matrix by a diagonal one, so that matrix inversion becomes trivial, even with nonuniform rather than only uniform sampling that was considered in earlier work. This approximation reduces the computational complexity of the estimator and estimation bound significantly. We numerically validate this approximation and compare our approach with the Maximum Likelihood Estimator (MLE) and Cramer-Rao Lower Bound (CRLB) for multivariate Gaussian distributions. Simulations show that our approach differs less than 0.1% from this MLE and CRLB when the observation time is large compared to the correlation time. Additionally, simulations show that in case of non-uniform sampling, we increase the performance in comparison to earlier work by an order of magnitude. We limit this study to correlated signals in the time domain, but the results are also applicable in the space domain.

In Moving Horizon Estimation (MHE) the computed estimate is found by solving a constrained finite-time optimal estimation problem in real-time at each sample in a receding horizon fashion. The constrained estimation problem can be solved by, e.g., interior-point (IP) or active-set (AS) methods, where the main computational effort in both methods is known to be the computation of the search direction, i.e., the Newton step. This is often done using generic sparsity exploiting algorithms or serial Riccati recursions, but as parallel hardware is becoming more commonly available the need for parallel algorithms for computing the Newton step is increasing. In this paper a newly developed tailored, non-iterative parallel algorithm for computing the Newton step using the Riccati recursion for Model Predictive Control (MPC) is extended to MHE problems. The algorithm exploits the special structure of the Karush-Kuhn-Tucker system for the optimal estimation problem. As a result it is possible to obtain logarithmic complexity growth in the estimation horizon length, which can be used to reduce the computation time for IP and AS methods when applied to what is today considered as challenging estimation problems. Furthermore, promising numerical results have been obtained using an ANSI-C implementation of the proposed algorithm, which uses Message Passing Interface (MPI) together with InfiniBand and is executed on true parallel hardware. Beyond MHE, due to similarities in the problem structure, the algorithm can be applied to various forms of on-line and off-line smoothing problems.

The speed of a wheeled vehicle is usually estimatedusing wheel speed sensors (WSS) or GPS. If these signals are unavailable, other methods must be used. We propose a novelapproach exploiting the fact that vibrations from rotating axles,with fundamental frequency proportional to vehicle speed, aretransmitted via the vehicle chassis. Using an accelerometer, these vibrations can be tracked to estimate vehicle speed whileother sources of vibrations act as disturbances. A state-space model for the dynamics of the harmonics is presented andformulated such that there is a conditional linear-Gaussiansubstructure, enabling efficient Rao-Blackwellized methods. Avariant of the Rao-Blackwellized point-mass filter is derived, significantly reducing computational complexity, and reducingthe memory requirements from quadratic to linear in thenumber of grid points. It is applied to experimental data from the sensor cluster of a car and validated using therotational frequency from WSS data. The proposed methodshows improved performance and robustness in comparisonto a Rao-Blackwellized particle filter implementation and afrequency spectrum maximization method.

A first course in Automatic control is presented.  A main objective of the course is to put most of the emphasis on the Implement and Operate phases in the process of developing a control system for a process. The course is built around a large amount of student active learning based on three extensive laboratory exercises, where each laboratory exercise can have duration of up to two weeks. For each of the laboratory exercises there is a sequence of learning activities supporting the students’ learning: Introductory lecture, problem solving session, preparation work, help-desk session, independent work in the laboratory, and a final demonstration of the control system. In addition there is a small project where the task is to write a manual for a process operator. The laboratory tasks involve implementation of a control system in an industrial PLC (Programmable Logic Controller) and development of an operator interface.

Parametric Cramer-Rao lower bounds (CRLBs) are given for discrete-time systems with non-zero process noise. Recursive expressions for the conditional bias and mean-square-error (MSE) (given a specific state sequence) are obtained for Kalman filter estimating the states of a linear Gaussian system. It is discussed that Kalman filter is conditionally biased with a non-zero process noise realization in the given state sequence. Recursive parametric CRLBs are obtained for biased estimators for linear state estimators of linear Gaussian systems. Simulation studies are conducted where it is shown that Kalman filter is not an efficient estimator in a conditional sense.

This paper investigates the usefulness of multi-frequency received signal strength (RSS) for indoor localization. Acollected set of data from four sites containing 7 frequencies fromdual receivers and a high accuracy reference positioning systemis presented. The collected data is also made publicly availablethrough ResearchGate. The data is analyzed with respect tospatial variations using Gaussian processes ( GP ). The resultsshow that there are more rapid signal variations across corridorsthan along them. The uniqueness of RSS fingerprints is analyzedsuggesting that sequences of measurements in smoothing, orsmoothing-like, algorithms that can handle temporary positionambiguities are likely the best choice for localization applications.

In this paper, we propose a new method to identify biochemical reaction networks (i.e. both reactions and kinetic parameters) from heterogeneous datasets. Such datasets can contain (a) data from several replicates of an experiment performed on a biological system; (b) data measured from a biochemical network subjected to different experimental conditions, for example, changes/perturbations in biological inductions, temperature, gene knock-out, gene over-expression, etc. Simultaneous integration of various datasets to perform system identification has the potential to avoid non-identifiability issues typically arising when only single datasets are used.

A skew-t variational Bayes filter (STVBF) is applied to indoor positioning with time-of-arrival (TOA) based distance measurements and pedestrian dead reckoning (PDR). The proposed filter accommodates large positive outliers caused by occasional non-line-of-sight (NLOS) conditions by using a skew-t model of measurement errors. Real-data tests using the fusion of inertial sensors based PDR and ultra-wideband based TOA ranging show that the STVBF clearly outperforms the extended Kalman filter (EKF) in positioning accuracy with the computational complexity about three times that of the EKF.

Gaussian process regression is a popular method for non-parametric probabilistic modeling of functions. The Gaussian process prior is characterized by so-called hyperparameters, which often have a large influence on the posterior model and can be difficult to tune. This work provides a method for numerical marginalization of the hyperparameters, relying on the rigorous framework of sequential Monte Carlo. Our method is well suited for online problems, and we demonstrate its ability to handle real-world problems with several dimensions and compare it to other marginalization methods. We also conclude that our proposed method is a competitive alternative to the commonly used point estimates maximizing the likelihood, both in terms of computational load and its ability to handle multimodal posteriors.

In order to investigate the cases in which an externally positive discrete-time system fails to have a minimal positive realization, in this paper we introduce the notion of minimal eventually positive realization, fr which the state update matrix becomes positive after a certain power. This property captures the idea that in the impulse response of an externally positive system the state of a minimal realization may fail to be positive, but only transiently. It is shown in the paper that whenever a minimal eventually positive realization exists, then the sequence of Markov parameters of the impulse response admits decimated subsequences for which minimal positive realizations exist and can be obtained by downsampling the eventually positive realization.

In Model Predictive Control (MPC) the control signal is computed by solving a constrained finite-time optimal control (CFTOC) problem at each sample in the control loop. The CFTOC problem can be solved by, e.g., interior-point or active-set methods, where the main computational effort in both methods is known to be the computation of the search direction, i.e., the Newton step. This is often done using generic sparsity exploiting algorithms or serial Riccati recursions, but as parallel hardware is becoming more commonly available the need for parallel algorithms for computing the Newton step is increasing. In this paper a tailored, non-iterative parallel algorithm for computing the Newton step using the Riccati recursion is presented. The algorithm exploits the special structure of the Karush-Kuhn-Tucker system for a CFTOC problem. As a result it is possible to obtain logarithmic complexity growth in the prediction horizon length, which can be used to reduce the computation time for popular state-of-the-art MPC algorithms when applied to what is today considered as challenging control problems.

One of the key challenges in identifying nonlinear and possibly non-Gaussian state space models (SSMs) is the intractability of estimating the system state. Sequential Monte Carlo (SMC) methods, such as the particle filter (introduced more than two decades ago), provide numerical solutions to the nonlinear state estimation problems arising in SSMs. When combined with additional identification techniques, these algorithms provide solid solutions to the nonlinear system identification problem. We describe two general strategies for creating such combinations and discuss why SMC is a natural tool for implementing these strategies.

Simultaneous localization and mapping (SLAM) is a well-known positioning approach in GPS-denied environments such as urban canyons and inside buildings. Autonomous/aided target detection and recognition (ATR) is commonly used in military application to detect threats and targets in outdoor environments. This papers present approaches to combine SLAM with ATR in ways that compensate for the drawbacks in each method. The methods use physical objects that are recognizable by ATR as unambiguous features in SLAM, while SLAM provides the ATR with better position estimates. Landmarks in the form of 3D point features based on normal aligned radial features (NARF) are used in conjunction with identified objects and 3D object models that replace landmarks when possible. This leads to a more compact map representation with fewer landmarks, which partly compensates for the introduced cost of the ATR. We analyze three approaches to combine SLAM and 3D-data; point-point matching ignoring NARF features, point-point matching using the set of points that are selected by NARF feature analysis, and matching of NARF features using nearest neighbor analysis. The first two approaches are is similar to the common iterative closest point (ICP). We propose an algorithm that combines EKF-SLAM and ATR based on rectangle estimation. The intended application is to improve the positioning of a first responder moving through an indoor environment, where the map offers localization and simultaneously helps locate people, furniture and potentially dangerous objects such as gas canisters.

We propose nested sequential Monte Carlo (NSMC), a methodology to sample from sequences of probability distributions, even where the random variables are high-dimensional. NSMC generalises the SMC framework by requiring only approximate, properly weighted, samples from the SMC proposal distribution, while still resulting in a correct SMC algorithm. Furthermore, NSMC can in itself be used to produce such properly weighted samples. Consequently, one NSMC sampler can be used to construct an efficient high-dimensional proposal distribution for another NSMC sampler, and this nesting of the algorithm can be done to an arbitrary degree. This allows us to consider complex and high-dimensional models using SMC. We show results that motivate the efficacy of our approach on several filtering problems with dimensions in the order of 100 to 1 000.

Data-efficient learning in continuous state-action spaces using very high-dimensional observations remains a key challenge in developing fully autonomous systems. In this paper, we consider one instance of this challenge, the pixels to torques problem, where an agent must learn a closed-loop control policy from pixel information only. We introduce a data-efficient, model-based reinforcement learning algorithm that learns such a closed-loop policy directly from pixel information. The key ingredient is a deep dynamical model that uses deep auto-encoders to learn a low-dimensional embedding of images jointly with a predictive model in this low-dimensional feature space. Joint learning ensures that not only static but also dynamic properties of the data are accounted for. This is crucial for long-term predictions, which lie at the core of the adaptive model predictive control strategy that we use for closed-loop control. Compared to state-of-the-art reinforcement learning methods for continuous states and actions, our approach learns quickly, scales to high-dimensional state spaces and is an important step toward fully autonomous learning from pixels to torques.

This paper concerns the estimation of approximate (linear) inverse models of block-oriented systems and the presented results give an improved understanding of these approximations. The estimated inverse is intended to be used as a pre- or postdistorter of the original system and a good inverse model would thus be one that, when used in series with the original system, produces a signal that resembles the original input. An inverse model of a nonlinear system can either be estimated in the standard way (from input to output) and then inverted, or directly (from output to input). This choice will affect the model. In the general case, the two modeling approaches will lead to different models, which will be shown for Hammerstein and Wiener systems. However, for a noise-free Hammerstein system with a white Gaussian input, the two approaches will result in the same model, up to a constant. When the two models are not equal, and the goal is to use the inverse as described above, it can be beneficial to estimate an approximate inverse directly. It will also be illustrated in an example how the inverse estimate can be used to get a nonparametric estimate of the nonlinearity in a block-oriented system. 

This paper presents a method for global pose estimation using inertial sensors, monocular vision, and ultra wide band (UWB) sensors. It is demonstrated that the complementary characteristics of these sensors can be exploited to provide improved global pose estimates, without requiring the introduction of any visible infrastructure, such as fiducial markers. Instead, natural landmarks are jointly estimated with the pose of the platform using a simultaneous localization and mapping framework, supported by a small number of easy-to-hide UWB beacons with known positions. The method is evaluated with data from a controlled indoor experiment with high precision ground truth. The results show the benefit of the suggested sensor combination and suggest directions for further work.

The well-known cooperative localization algorithm, ‘sum-product algorithm over a wireless network’ (SPAWN) hastwo major shortcomings, a relatively high computationalcomplexity and a large communication load. Using the Gaus-sian mixture model with a model selection criterion and thesigma-point (SP) methods, we propose the SPAWN-SP toovercome these problems. The SPAWN-SP easily accommo-dates different localization scenarios due to its high flexibilityin message representation. Furthermore, harsh LOS/NLOSenvironments are considered for the evaluation of coopera-tive localization algorithms. Our simulation results indicatethat the proposed SPAWN-SP demonstrates high localizationaccuracy in different localization scenarios, thanks to its highflexibility in message representation.

A marginal version of the Weiss-Weinstein bound (WWB) is proposed for discrete-time nonlinear filtering. The proposed bound is calculated analytically for linear Gaussian systems and approximately for nonlinear systems using a particle filtering scheme. Via simulation studies, it is shown that the marginal bounds are tighter than their joint counterparts.

Positioning in radio networks is a well establishedresearch area. The dominating approach has been that positioningalgorithms are implemented in the higher levels of the communicationsystem based on position related information derivedin the lowest (physical) layer. Examples of measurement includereceived signal strength (RSS), time of arrival (TOA), angleof arrival (AOA), and fusion and filtering is a straightforwardtask. The technical driver for positioning has been E911 andfor commercially driver comes from location based services andlogistics management. These demands are fundamental in thedevelopment of positioning in future radio networks standards.There is today a trend for accuracy demand that goes beyondwhat can be achieved with todays measurements. Another trendis that measurements and positioning algorithms are approachingeach other, so some parts of the positioning are performed on thechip-sets (lowest layer) and low-level measurements are availableto the operating system (highest level). The purpose of thissurvey is to describe this trend in more detail, with examples ofdevelopments in cellular networks as well as WiFi and Bluetooth.

A ship's roll dynamics is sensitive to the mass and mass distribution. Changes in these physical properties might introduce unpredictable behavior of the ship and a worst-case scenario is that the ship will capsize. In this paper, a recently proposed approach for online estimation of mass and center of mass is validated using experimental data. The experiments were performed using a scale model of a ship in a wave basin. The data were collected in free run experiments where the rudder angle was recorded and the ship's motion was measured using an inertial measurement unit. The motion measurements are used in conjunction with a model of the roll dynamics to estimate the desired properties. The estimator uses the rudder angle measurements together with an instrumental variable method to mitigate the influence of disturbances. The experimental study shows that the properties can be estimated with quite good accuracy but that variance and robustness properties can be improved further.

In this paper, a semi-parametric model for RSS measurements is introduced that can be used to predict coverage in cellular radio networks. The model is composed of an empirical log-distance model and a deterministic antenna gain model that accounts for possible non-uniform base station antenna radiation. A least-squares estimator is proposed to jointly estimate the path loss and antenna gain model parameters. Simulation as well as experimental results verify the efficacy of this approach. The method can provide improved accuracy compared to conventional path loss based estimation methods. 

The extended Kalman filter (EKF) has been animportant tool for state estimation of nonlinear systems sinceits introduction. However, the EKF does not possess the same optimality properties as the Kalman filter, and may perform poorly. By viewing the EKF as an optimization problem it is possible to, in many cases, improve its performance and robustness. The paper derives three variations of the EKF by applying different optimisation algorithms to the EKF costfunction and relate these to the iterated EKF. The derived filters are evaluated in two simulation studies which exemplify the presented filters.

Visual animal tracking is a challenging problem generally requiring extended target models, group tracking and handling of clutter and missed detections. Furthermore, the dolphin tracking problem we consider includes basin constraints, shadows, limited field of view and rapidly changing light conditions. We describe the whole pipeline of a solution based on a ceiling-mounted fisheye camera that includes foreground segmentation and observation extraction in each image, followed by a target tracking framework. A novel contribution is a potential field model of the basin edges as a part of the motion model, that provides a robust prediction of the dolphin trajectories in phases with long segments of missed detections. The overall performance on real data is quite promising.

Subspace identification is revisited in the scope of nuclear norm minimization methods. It is shown that essential structural knowledge about the unknown data matrices in the data equation that relates Hankel matrices constructed from input and output data can be used in the first step of the numerical solution presented. The structural knowledge comprises the low rank property of a matrix that is the product of the extended observability matrix and the state sequence and the Toeplitz structure of the matrix of Markov parameters (of the system in innovation form). The new subspace identification method is referred to as the N2SID (twice the N of Nuclear Norm and SID for Subspace IDentification) method. In addition to include key structural knowledge in the solution it integrates the subspace calculation with minimization of a classical prediction error cost function. The nuclear norm relaxation enables us to perform such integration while preserving convexity. The advantages of N2SID are demonstrated in a numerical open- and closed-loop simulation study. Here a comparison is made with another widely used SID method, i.e. N4SID. The comparison focusses on the identification with short data batches, i.e. where the number of measurements is a small multiple of the system order.

We are concerned with the problem of detecting an overtaking vehicle using a single camera mounted behind the ego-vehicle windscreen. The proposed solution makes use of 1D optical flow evaluated along lines parallel to the motion of the overtaking vehicles. The 1D optical flow is computed by tracking features along these lines. Based on these features, the position of the overtaking vehicle can also be estimated. The proposed solution has been implemented and tested in real time with promising results. The video data was recorded during test drives in normal traffic conditions in Sweden and Germany.

In vehicle load detection and estimation, having an accurate estimate of the height of the Center of Gravity (CG) and the Roll Center (RC) is crucial. In this paper, taking advantage of the availability of Inertial Measurement Units (IMUs) in automotive vehicles, the nominal CG and RC heights are estimated and used for the main purpose of roof load detection. RC estimation is formulated as an Errors-in-Variables regression model. The Total and Corrected Least Squares (TLS and CLS) approaches are applied and a comparison is performed. Using the knowledge of the covariance matrix, CLS provides consistent estimates of the RC height. The estimated RC and CG heights are used to estimate the roof load using a greybox model and the method is applied to measurements from normal driving conditions.

Common data preprocessing routines often introduce considerable flaws in laser-based tracking of extended objects. As an alternative, extended target tracking methods, such as the Gamma-Gaussian-Inverse Wishart (GGIW) probability hypothesis density (PHD) filter, work directly on raw data. In this paper, the GGIW-PHD filter is applied to real world traffic scenarios. To cope with the large amount of data, a mixture clustering approach which reduces the combinatorial complexity and computation time is proposed. The effective segmentation of raw measurements with respect to spatial distribution and motion is demonstrated and evaluated on two different applications: pedestrian tracking from a vehicle and intersection surveillance.

Enumerative nonlinear model predictive control for speed tracking problem of linear induction motors has been presented in [1], where the authors show that this control scheme has better performance as compared to direct torque control. In this paper, the authors show that using a load observer for integral action, the performance can be further improved. Specifically simulation results show that a load observer results in better tracking properties and offers more robust control.

We propose an improved proposal distribution in the Particle Metropolis-Hastings (PMH) algorithm for Bayesian parameter inference in nonlinear state space models. This proposal incorporates second-order information about the parameter posterior distribution, which can be extracted from the particle filter already used within the PMH algorithm. The added information makes the proposal scale-invariant, simpler to tune and can possibly also shorten the burn-in phase. The proposed algorithm has a computational cost which is proportional to the number of particles, i.e. the same as the original marginal PMH algorithm. Finally, we provide two numerical examples that illustrates some of the possible benefits of adding the second-order information.

We derive a new Sequential-Monte-Carlo-based algorithm to estimate the capacity of two-dimensional channel models. The focus is on computing the noiseless capacity of the 2-D (1, ∞) run-length limited constrained channel, but the underlying idea is generally applicable. The proposed algorithm is profiled against a state-of-the-art method, yielding more than an order of magnitude improvement in estimation accuracy for a given computation time.

Stochastic dynamical systems are fundamental in state estimation, systemidentification and control. System models are often provided incontinuous time, while a major part of the applied theory is developedfor discrete-time systems. Discretization of continuous-time models ishence fundamental. We present a novel algorithm using a  combination of Lyapunov equations and analytical solutions, enabling  efficient implementation in software. The proposed method  circumvents numerical problems exhibited by standard algorithms in  the literature. Both theoretical and simulation results are  provided.

Magnetometers and inertial sensors (accelerometers and gyroscopes) are widely used to estimate 3D orientation. For the orientation estimates to be accurate, the sensor axes need to be aligned and the magnetometer needs to be calibrated for sensor errors and for the presence of magnetic disturbances. In this work we use a grey-box system identification approach to compute maximum likelihood estimates of the calibration parameters. An experiment where the magnetometer data is highly disturbed shows that the algorithm works well on real data, providing good calibration results and improved heading estimates. We also provide an identifiability analysis to understand how much rotation is needed to be able to solve the calibration problem.

In this paper, a method for estimating physical parameters using limited sensors is investigated. As a case study, measurements from an IMU are used for estimating the change in mass and the change in center of mass of a ship. The roll motion is studied and an instrumental variable method estimating the parameters of a transfer function from the tangential acceleration to the angular velocity is presented. It is shown that only a subset of the unknown parameters are identifiable simultaneously. A multi-stage identification approach is presented as a remedy for this. A limited simulation study is also presented to show the properties of the estimator. This shows that the method is indeed promising but that more work is needed to reduce the variance of the estimator.

There is today no established automated method for testing vehicles or tyres, and the most common option is using professional drivers for this purpose. The tests are supposed to be fair and repeatable, which means using human drivers for these kinds of vehicle testing is not an option. Using a steering robot modelled to drive as a human is therefore preferable. The approach described in this paper shows how a driver model can be created by using a control algorithm based on gathered data from human drivers performing double lane change (DLC) manoeuvres in a simulator. The implemented controller shows how human drivers’ behaviors can be captured using control theory.

Large-scale interconnected uncertain systems commonly have large state and uncertainty dimensions. Aside from the heavy computational cost of solving centralized robust stability analysis techniques, privacy requirements in the network can also introduce further issues. In this paper, we utilize IQC analysis for analyzing large-scale interconnected uncertain systems and we evade these issues by describing a decomposition scheme that is based on the interconnection structure of the system. This scheme is based on the so-called chordal decomposition and does not add any conservativeness to the analysis approach. The decomposed problem can be solved using distributed computational algorithms without the need for a centralized computational unit. We further discuss the merits of the proposed analysis approach using a numerical experiment.