基于修正导航向量场的AUV自主避障方法  

Autonomous Obstacle Avoidance for AUV Based on Modified Guidance Vector Field

作　　者：姚鹏 解则晓[1] YAO Peng;XIE Ze-Xiao(College of Engineering,Ocean University of China,Qingdao 266100)

机构地区：[1]中国海洋大学工程学院,青岛266100

出　　处：《自动化学报》2020年第8期1670-1680,共11页Acta Automatica Sinica

基　　金：山东省自然科学基金(ZR2018BF016);中国博士后科学基金(2017M622278);中央高校基本科研业务费(201713046)资助。

摘　　要：针对复杂海洋环境下的自治水下机器人(Autonomous underwater vehicle,AUV)三维避障问题,本文提出了一种高效的修正导航向量场方法.构建自由空间下的初始导航向量场,引导AUV以最短路径向目标点航行.定义修正矩阵来量化描述障碍物对初始导航向量场的影响,得到障碍空间下的修正导航向量场,使得AUV向目标点航行的同时躲避静态障碍.通过结合障碍物运动速度,分别构建相对初始导航向量场与相对修正导航向量场,并采取有限时域推演与调整策略,最终引导AUV安全躲避动态障碍.仿真结果表明,本方法能较好地应用于复杂海洋环境下的AUV避障任务.An efficient method called the modified guidance vector field is proposed to solve the three-dimensional obstacle avoidance problem for autonomous underwater vehicle(AUV)in complex ocean environment.The initial guidance vector field in free space is first constructed to guide AUV to the destination along the shortest path.Then the modulation matrix is defined to quantify the influence of obstacles on the initial guidance vector field,and the modified guidance vector field in obstacle space is hence obtained,where AUV will avoid static obstacles when navigating to the destination.The referred velocity of dynamic obstacles is introduced to construct the relative initial/modified guidance vector field,and the limited time domain based derivation and adjustment strategy is also utilized to guide AUV avoiding dynamic obstacles safely.Finally the simulation results demonstrate that this method applies to the obstacle avoidance mission for AUV well in complex ocean environment.

关 键 词：自治水下机器人 三维避障 初始导航向量场 修正导航向量场 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置]