基于行为策略的AUV全覆盖信度函数路径规划算法  

Complete Coverage Belief Function Path Planning Algorithm of Autonomous Underwater Vehicle Based on Behavior Strategy

作　　者：甘文洋 朱大奇[1] Gan Wenyang;Zhu Daqi(Laboratory of Underwater Vehicles and Intelligent Systems, Shanghai 201306, China.)

机构地区：[1]上海海事大学水下机器人与智能系统实验室,上海201306

出　　处：《系统仿真学报》2018年第5期1857-1868,共12页Journal of System Simulation

基　　金：国家自然科学基金(51279098;51575336);上海市科委创新行动计划(13510721400)

摘　　要：针对自治水下机器人全覆盖路径规划问题,通过引入行为策略,结合地图信度函数概念,提出一种基于行为策略的全覆盖信度函数路径规划算法,使AUV完成全覆盖任务的同时自动避开障碍物。根据环境信息构建栅格信度函数,用不同的函数值区分障碍物、已覆盖和未覆盖栅格;AUV根据路径规划策略选择下一步航行位置,若AUV航行到障碍物边缘,则利用行为策略选择下一步航行位置,否则使用栅格信度函数选择下一步航行位置。通过二维和三维仿真实验证明了所提算法不仅能够完成全覆盖任务,同时减少了AUV陷入死区的次数,降低了遍历重复率。For the complete coverage path planning of autonomous underwater vehicle （AUV）, a complete coverage belief function path planning algorithm based on behavior strategy is proposed by introducing the behavior strategy and using the concept of map belief function, which can make AUV accomplish the complete coverage and avoid obstacles automatically. The grid belief function is constructed based on environmental information, using different function values to distinguish between the obstacle, the covered and the uncovered grid. AUV selects the next navigation position by path planning strategy. Next navigation position is selected by behavior strategy if AUV navigates to the edge of obstacle. Otherwise, belief function is used to select navigation position of AUV. By simulation experiments in 2-D and 3-D environment, the algorithms mentioned in this paper is proved to be capable of accomplishing the complete coverage, decreasing the number of AUV into dead zone and reducing the overlay repetition rate.

关 键 词：自治水下机器人 全覆盖路径规划 信度函数 行为策略 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]