三维水下环境中的多AUV围捕路径规划算法  

Multi-AUV hunting path planning in 3D underwater environment

作　　者：蒋骁迪 朱大奇[1] 陈铭治[1] 甘文洋 JIANG Xiaodi;ZHU Daqi;CHEN Mingzhi;GAN Wenyang(Shanghai Engineering Research Center of Intelligent Maritime Search&Rescue and Underwater Vehicles,Shanghai Maritime University,Shanghai 201306,China)

机构地区：[1]上海海事大学智能海事搜救与水下机器人上海工程技术研究中心,上海201306

出　　处：《现代电子技术》2021年第11期155-159,共5页Modern Electronics Technique

基　　金：国家自然科学基金重大研究计划(91748117)。

摘　　要：受到三维水下环境覆盖范围广和水下机器人(AUV)能源有限等现实条件的制约,单个AUV往往很难完成作业规模较大的水下任务,故提出一种基于位置分配和Glasius生物启发神经网络(GBNN)的围捕占位路径规划算法,来解决三维水下环境中的多AUV围捕路径规划问题。首先,根据围捕AUV到达逃逸AUV的围捕总距离最短原则分配围捕点,确保围捕AUV与围捕点均衡分配并避免路径冲突;接着,各围捕AUV根据所分配的围捕点,利用改进的生物启发神经网络的目标引导与自适应避障功能,规划各个围捕AUV到达逃逸AUV周围围捕点的路径,完成围捕任务。仿真实验验证所提算法能够在三维水下静态障碍物环境中完成高效的协作围捕任务。Due to the limitations of the wide coverage of the 3D underwater environment and the limited energy for autonomous underwater vehicle(AUV),it is often difficult for a single AUV to complete a wide range of underwater tasks.Therefore,a hunting occupation path planning algorithm based on location allocation and Glasius bio-inspired neural network is proposed for the multi-AUV hunting path planning in 3D underwater environment.First,according to the principle of the shortest total distance between the hunting AUV and the escaped AUV,the capture point around the escaped AUV is allocated to ensure the balanced distribution and avoid the path conflict.Then,the path of each hunting AUV to the capture point around the escape AUV is planned by the goal guidance and adaptive obstacle avoidance function of the Glasius bio-inspired neural network,so as to complete the round-up task.The simulation results show that the proposed algorithm can achieve efficient cooperative hunting task in 3D underwater static obstacle environment.

关 键 词：水下机器人 多AUV协作围捕 围捕占位 路径冲突 位置分配 路径规划 避障 生物启发神经网络 

分 类 号：TN911.1-34[电子电信—通信与信息系统] TP242.6[电子电信—信息与通信工程]