1P1-L04 Time-Space Path Planning Algorithm for Mobile Robots in Dynamic Environments Based on Probabilistic Sampling(Wheeled Robot/Tracked Vehicle)
Yuki Ishikawa, Jun Miura · The Proceedings of JSME annual Conference on Robotics and Mechatronics (Robomec) · 2011
This paper describes time-space path planning algorithms for mobile robots. In usual environments, both static obstacles, like walls and pillars, and dynamic ones, like people exist that prevent a robot from going directly to a destination. Our path planner generates paths considering kinematic and dynamic constraints of actual robots. For efficiency, the planner uses a random search method based on a probabilistic sampling. To avoid both static and dynamic obstacles effectively, we compute a time-space potential field based on a short-term obstacle motion prediction, and then use it as a bias (time-space goal-directed bias) for sampling. A time-space goal-directed bias gives larger weights to shorter and safer paths towards a destination. We evaluate the proposed algorithm in a simulated environment.