Application of Time Dependent Probabilistic Collision State Checkers in Highly Dynamic Environments
This paper proposes a motion planning method for a mobile robot in the situation where there are both static and moving obstacles. If the robot cannot communicate with moving obstacles, it has to predict their future movement in order to plan the safe and e cient motion. Since such a prediction includes uncertainty, the proposed method explicitly considers the uncertainty in motion planning. We use a probabilistic model of the uncertainty and select the motion which minimizes the expected time of reaching the destination. We also utilize the knowledge of possible paths of moving obstacles, which is applicable to usual structured environments. Simulation results validate the e ectiveness of the method.