Probability-based Path Planning for Stochastic Nonholonomic Systems with Obstacle Avoidance
Author | : Jianping Lin |
Publisher | : |
Total Pages | : 106 |
Release | : 2015 |
ISBN-10 | : OCLC:922552755 |
ISBN-13 | : |
Rating | : 4/5 ( Downloads) |
Download or read book Probability-based Path Planning for Stochastic Nonholonomic Systems with Obstacle Avoidance written by Jianping Lin and published by . This book was released on 2015 with total page 106 pages. Available in PDF, EPUB and Kindle. Book excerpt: There exist various path planning methods in robotics. The Probabilistic Roadmap and Rapidly-exploring Random Tree (RRT) became popular in recent decades. It is known that the RRT is more suitable for nonholonomic systems. The RRT is a sampling-based algorithm which is designed for path planning problem and is efficient to handle high-dimensional configuration space (C-space) and nonholonomic constraints. Under the constraints, the RRT can generate paths between an initial state and a goal state while avoiding obstacles. However it does not guarantee that the resulting path is optimal. In systems with stochasticity, targeting error and closeness of the obstacle to the planned path can be considered to obtain the optimal path. In this thesis, the targeting error is defined as the root-mean-square (RMS) distance from the path samples to the desired target and the closeness is defined as probability of obstacle collision. Then, a cost function is defined as a sum of the targeting error and the obstacle closeness, and numerically minimized to find the path. The RRT result serves as an initial starting point for this subsequent optimization.