基于协同进化的多AUV多区域覆盖路径规划  

Multi-AUV multi-regional coverage path planning based on coevolution

作　　者：蔡畅 陈丹 蔡磊[2] Cai Chang;Chen Dan;Cai Lei(School of Information and Science Technology,Shijiazhuang Tiedao University,Shijiazhuang 050043,China;School of Automation,Northwestern Polytechnical University,Xi’an 710072,China)

机构地区：[1]石家庄铁道大学信息科学与技术学院,河北石家庄050043 [2]西北工业大学自动化学院,陕西西安710072

出　　处：《海洋学报》2024年第11期115-124,共10页

基　　金：国家自然科学基金(62071383)。

摘　　要：针对多自主水下航行器(Autonomous Underwater Vehicle, AUV)水下覆盖任务过程中的突发情况,研究了多AUV的覆盖路径重规划问题,提出了一种多机器人-多区域覆盖路径规划(Multi-robot Multi-regional Coverage Path Planning, M^(2)CPP)方法,为可用AUV重新分配未覆盖区域并规划覆盖路径。首先,通过割草机算法确定每个区域中的内部路径和候选入口位置。然后,采用协同进化方法求解最优的区域分配、区域顺序及各区域的最优入口,3个种群协同进化,共同决定所有AUV的完整路径,保证种群多样性,避免陷入局部最优。仿真结果表明,本文方法在根据初始位置和剩余能量为多AUV重规划较短路径的基础上,优化路径结构,保证多AUV工作量均衡,可以较好解决该背景下的路径重规划问题。In response to contingencies that arise during the underwater coverage missions of multiple autonomous underwater vehicles(AUVs),this study addresses the problem of coverage path replanning for multiple AUVs.A multi-robot multi-regional coverage path planning(M^(2)CPP)method is proposed to reassign uncovered areas to available AUVs and plan their coverage paths.Initially,the lawnmower algorithm is employed to determine the internal paths and candidate entry points within each region.Subsequently,a coevolutionary approach is utilized to solve for the optimal region allocation,region sequence,and the best entry points for each region.Three populations coevolve collaboratively to determine the complete paths for all AUVs,ensuring population diversity and preventing convergence into local optima.Simulation results demonstrate that the proposed method not only replans shorter paths for multiple AUVs based on their initial positions and remaining energy but also optimizes the path structure to ensure a balanced workload among the AUVs,effectively resolving the replanning issue under such scenarios.

关 键 词：多机器人多区域覆盖路径规划 覆盖路径重规划 多AUV系统 协同进化 任务分配 

分 类 号：TP249[自动化与计算机技术—检测技术与自动化装置]