Global Trajectory Planning Based on DWMR Dynamics in Circular C-space

Abstract

This paper proposes an efficient near time-optimal trajectory planning algorithm for differential wheeled mobile robots(DWMRs) in a circular C-space under constraints on robot's kinematics, dynamics, and motor's voltage. This problem is considered to be complex if there exist lots of obstacles to pass and robot's dynamics including motor's dynamics should be considered. A* algorithm is incorporated into the modified TOTP (Time Optimal Trajectory Planning) algorithm [1] to find near minimum time trajectory without any homotopy class by defining a node as an obstacle with direction information. In addition, TOTP algorithm is improved so that the proposed algorithm can be applied to the environment without any assumption about the distance between obstacles. Simulations show that the proposed algorithm is applied in an environment where multiple obstacles exist. An experiment shows that the generated trajectory by the proposed algorithm is well followed by the robot.