为提高机器鱼的水下路径规划效率,更好地完成水下工作,提出一种基于改进双向快速搜索随机树(bidirectional rapidly-exploring random trees,Bi-RRT)算法的机器鱼路径规划方法.以研制的混合驱动机器鱼为研究对象,介绍其结构模型和运动控制模式,为后续试验验证提供物理样机.针对Bi-RRT算法存在的采样随机、路径冗余、效率不高等问题,融合生长引导机制和连接强化机制改进Bi-RRT算法,加入生长引导机制,改善随机树生长随机、两树连接慢的问题;加入连接强化机制提高算法搜索速度.对搜索路径进行优化处理,通过剔除冗余节点、插入优化节点,改善路径质量,对路径进行平滑处理,使路径更适合机器鱼航行,实现机器鱼路径规划任务.仿真结果表明,与传统Bi-RRT算法及其他衍生快速搜索随机树(rapidly-exploring random tree,RRT)算法相比,改进的Bi-RRT算法相较于改进前节点数减少约 50.8%,路径长度缩短约 19%,搜索时间减少约 65.3%.
Robot fish path planning method based on improved Bi-RRT algorithm
In order to improve the efficiency of underwater path planning for robotic fish and better complete underwater work,a path planning method for robotic fish based on improved bidirectional rapidly-exploring random trees(Bi-RRT)algorithm was pro-posed.A hybrid drive robotic fish was developed as the research object,and its structural model and motion control mode were intro-duced,which provided a physical prototype for subsequent experimental verification.In response to the problems of sampling ran-domness,path redundancy,and low efficiency in the Bi-RRT algorithm,a growth guidance mechanism and connection strengthening mechanism were integrated to improve the Bi-RRT algorithm,and a growth guidance mechanism was added to improve the problem of random tree growth and slow connection between two trees;added connection reinforcement mechanism to improve algorithm search speed.The search path was optimized by removing redundant nodes and inserting optimization nodes to im-prove path quality.The path was smoothed to make it more suitable for robot fish navigation and to achieve robot fish path planning tasks.The simulation results showed that compared with the traditional Bi-RRT algorithm and other derived rapidly-exploring random tree(RRT)algorithms,the improved Bi-RRT algorithm reduced the number of nodes by about 50.8%,the path length by about 19%,and the search time by about 65.3%compared to the previous one.
万方数据
维普
