video.mp4
A path planning and control method is developed for the scenario of a non-holonomic robot serving food in a restaurant. The path planning algorithm used is PRM*, where a semi-random point sampling algorithm is used. Graph search is performed by the Dijkstra algorithm, after which b-splines is used to smoothen the path. By using a pure pursuit model in combination with a PID-controller, the serving robot can follow the trajectory within predefined limits. Finally, the results are discussed, commenting on the performance and limitations of the system, followed by future recommendations. Take a look at video for a quick summary. Read the report provided for a detailed explanation of mathematics behind the code.
- Chinmay Polyaramesh 5135125
- Stan Zwinkels 4630726
- Jelmer de Wolde 4705041
math
numpy
scipy
PIL
random
os
pygame
Run the scipt Main.py. A PyGame windows will be opened. First run always takes a while, since borders
are getting plotted and graph gets sampled. The location can be selected using mouse click. Map can be
changed on line 32. Three maps are supplied (A, B and C).
