REAL-TIME PATH PLANNING OF MULTIPLE MOBILE ROBOTS IN A DYNAMIC CERTAIN ENVIRONMENT

摘要

We propose an algorithm for real-time path planning of multiple mobile robots in a dynamic certain environment. Dynamic certain environment means that shapes and velocities of the obstacles can be measured in real-time by the robot's sensory system. It is difficult to obtain real-time and completeness simultaneously for path planning of multiple robots without suffering from oscillation and deadlock, especially in a dynamic environment. In this paper, the harmonic potential field for multiple mobile robots is built by using the panel method known in fluid mechanics; attainability and security are investigated so that robots could reach goal positions while avoiding obstacles without any collision. A new velocity control strategy for multiple robots and a set of strategies for collision avoidance is proposed. Simulation and real robot experiment results indicate that it can achieve real-time and completeness simultaneously. The strategies for collision avoidance can be applied to avoid the oscillation and deadlock. The velocity control strategy is showed to be highly effective in reducing planning time.