About the project

This is a project in automatic control in a project course (Control Project Labortory, CDIO) held by the Department of Electrical Engineering (ISY) at Linköping University. The course is given to master students in automatic control.

The platform given includes a race track with radio controlled 1:43 scale cars and a truck, along with two infrared cameras and a video projector in the ceiling. This project is the fourth iteration of a system introduced in 2011 and will be used for educational and demonstration purposes as well as for reaserch.

Expansion 2014
This years addition is a further development of the system regarding autonomous control and obstacle detection. GPGO (Gaussian Processes for Global Optimization) has been implemented and is used to autonomously improve the laptimes. Virtual sensors have been implemented to the car and it's now possible to detect obstacles on the track. Rapidly exploring Random Tree (RRT) is another new implementation that calculates a new trajectory when the car detects an obstacle. This trajectory is then followed to avoid and pass the obstacle when it's possible. The truck can now be controlled from a computer and the the angles between the truck and the semitrailer can be estimated.

When driving without obstacles on the track, the control system autonomously calculates a new driving profile after each lap, in order to improve the lap time from the last lap. This is done with GPGO and since it was implemented in this years project, it has a huge potential for improvement.

When driving in presence of obstacles, the car follows a previously defined trajectory and stops when it detects an obstacle. This triggers the RRT program to calculate a new trajectory which is then followed by the car until the obstacle has been passed and the car once again follows the previously defined trajectory. The RRT algorithm approach is also a completely new method and has therefore big possibilities to improvement.

Linköpings Tekniska Högskola

12 hp

Course code / Course:
TSRT10 / Control Project Laboratory, CDIO

Daniel Axehill

Isak Nielsen

Kristoffer Lundahl