改进Bi-RRT算法的AGV全局路径规划

Global Path Planning for AGV Based on Improved Bi-RRT Algorithm

针对双向快速扩展随机树(Bidirectional Rapidly-Exploring Random Tree,Bi-RRT)算法中采样随机性大、路线贴合障碍物、生成路径存在大量冗余点等问题,提出了一种适用于自动导引运输车(Automatic Guided Vehicle,AGV)全局路径规划的改进Bi-RRT算法。该算法将基于概率p的目标偏向策略与A*算法中的代价估计思想结合起来,形成双重目标偏向策略以求得更优路径;在路径生成过程中加入安全距离判断机制,消除生成路径与障碍物贴合现象;此外,动态检测新节点与目标点的通路状态,减少不必要的随机采样过程;最后,利用遍历算法对所得路径进行后处理,滤除冗余点,使最终路径干净简洁。在MATLAB平台上对算法进行仿真实验,结果表明,改进算法相较于传统Bi-RRT,搜索时间更短,规划路径更优且可执行。

Aiming at problems such as large sampling randomness,route fitting obstacles,and numerous redundant points in generated path in Bidirectional Rapidly exploring Random Tree(Bi-RRT)algorithm,an improved Bi-RRT algorithm suitable for global path planning of Automatic Guided Vehicle(AGV)is proposed.In order to get a better path,the improved algorithm combines the target bias strategy based on probability p with the idea of cost estimation in A*algorithm to form a dual target bias strategy.Introduces a safe distance judgment mechanism in the path generation process to eliminate the phenomenon that the path fits the obstacles.Moreover,dynamically detects the path connectivity between the new node and the target point,reducing unnecessary random sampling processes.Finally,the path is post-processed by traversal algorithm to filter out redundant points and make the final path concise.Simulation experiments on the MATLAB platform show that compared with traditional Bi-RRT,the improved algorithm has shorter search time,a better and executable planned path.