Nonholonomic Motion Planning by Means of Particles

In this article a new approach to planning of a nonholonomic motion is presented. A flexible, intelligent planner based on a static map and the topology of the robot´s environment has been developed. The approach uses ´particles´ to construct automatically a path between two given locations. The generated path is a smooth trajectory, where the length of the path is kept at a minimum and obstacles are avoided. This concept applies to robots meeting the restrictions of a Dubin´s car (nonholonomic robot that can only move forward). After the basic concepts of the approach has been described, simulations will be presented