运动规划
采样(信号处理)
算法
计算机科学
随机树
随机性
移动机器人
数学优化
路径(计算)
启发式
概率逻辑
概率密度函数
数学
机器人
人工智能
统计
计算机视觉
滤波器(信号处理)
程序设计语言
标识
DOI:10.1109/auteee56487.2022.9994543
摘要
Bidirectional Fast Expanding Random Tree (Bi-RRT) algorithm has the problems of long search time and low sampling efficiency in path planning in complex environment due to the randomness of sampling points. Therefore, an improved Bi-RRT path planning algorithm for mobile robots was proposed;The algorithm introduces a heuristic search strategy, takes the starting point and the ending point of the robot as the center, constructs a two-dimensional Gaussian distribution function, and uses this probability density function to constrain the generation of sampling points, so that the spatial sampling points closer to the target point have a higher probability of occurrence, while retaining some uniformly distributed sampling points. In this way, the sampling process can not only make use of the location information of the target point but also ensure the probabilistic completeness of the algorithm; With the guidance of heuristic sampling points designed by the algorithm, two random trees can grow rapidly toward the target area, which reduces the blindness of the search and improves the efficiency of the search; Simulation results: Compared with the basic BI-RRT algorithm, the planning time of the improved algorithm is shortened by 43.9% in complex environment, the number of extended nodes is reduced by 41.4%, and the path length is optimized by 8.1%. The influence of the ratio of Gaussian distribution sampling points to the total number of sampling points on the performance of the algorithm is analyzed.
科研通智能强力驱动
Strongly Powered by AbleSci AI