摘要
Abstract
Aiming at the problems of path redundancy,low convergence efficiency,and poor adaptability to narrow environments of the traditional BI-RRT algorithm in the path planning of biomimetic robotic fish,a three-stage improvement strategy was proposed.Firstly,by dynamically adjusting the target bias probability and using an adaptive step size,combined with collision history feedback to reduce invalid nodes,the search efficiency was improved.Secondly,a greedy pruning technique was adopted to remove redundant nodes and shorten the path length.Finally,a cubic B-spline interpolation and a potential field guiding smoothing method were introduced to optimize the path continuity and avoid obstacle penetration.Simulation experiments showed that the improved algorithm exhibited significant advantages in the ordinary map environment.Compared with the traditional RRT algorithm,its path length was shortened by 26.6%,the planning time was reduced by 88.9%,and the number of nodes was only 23.7%of that of the RRT algorithm.When compared with the BI-RRT algorithm,the path length was shortened by 26.5%,the planning time was reduced by 59.7%,and the number of nodes was decreased by 47.5%.Moreover,the success rate of the algorithm remained 100%in the tests of this map.This algorithm improved the path quality and real-time performance,providing a reference solution for the motion planning of biomimetic robots in complex underwater environments.关键词
仿生机器鱼/路径优化/改进BI-RRT算法Key words
bionic robot fish/path optimization/improved BI-RRT algorithm分类
信息技术与安全科学
