Collision free path planning in case of moving obstacles

OData support
Supervisor:
Gincsainé Dr. Szádeczky-Kardoss Emese
Department of Control Engineering and Information Technology

Humanity is chasing automation in all areas of life ever since the Industrial Revolution. This process can be well observed in the continuous development and ever so wide application of robotic technology. In recent years, the development of autonomous vehicles has been stepping up.

The primary task of autonomous mobile robots is to reach from one place to another, while avoiding damage during movement either in themselves or in their environment. The robot needs to maneuver in a dynamic environment. The key to solving these problems is the motion planning algorithm.

The main purpose of my work is to present a motion planning algorithm that is able to avoid static and moving obstacles in the robot's environment. Based on that I used a Non-Linear Velocity Obstacles method, which is capable of determining the robot's trajectory and velocity profiles. The algorithm was implemented in Matlab.

The robot performs a series of evasive maneuvers based on the motion design, while trying to dodge all the obstacles. To choose the velocity of the robot I used two heuristic search methods (To Goal, Maximum Velocity). I have to examine their advantages and disadvantages.

To test the algorithm, I created a simulator, in which you can track the flow of motion planning during the movement of the robot and obstacles.

The possibility of the use of robots will grow in the future, so there are many opportunities to further develop the subject.

Downloads

Please sign in to download the files of this thesis.