Dynamic path planning for coordinated motion of multiple mobile robots

Abstract

This paper presents an approach of calculating coordinated paths for multiple mobile robots operating in the same environment. Therefore, the presented approach extends the configuration space with a time component. Based on an A∗ algorithm, it calculates collision free paths of multiple robots. Moreover, it is capable of replanning the path in dynamic environments without doing an iterated A∗ path planning from scratch. The correctness of the approach is proven, and experiments show the achieved improvement in computation time compared to an iterated path planning from scratch. A lattice type state space representation based on short motion sequences is applied to the approach, resulting in more realistic and traversable trajectories for vehicles. The feasibility is shown by letting a simulated convoy turn autonomously to the opposite direction.