大规模复杂变电站巡检机器人分层路径规划  

Hierarchical Motion Planning of Large-scale Complex Substation Inspection Robots

作　　者：李睿欣 汤明东 李海英 黄微 LI Rui-xin;TANG Ming-dong;LI Hai-ying;HUANG Wei(College of Mechanical Engineering,University of Shanghai for Science and Technology,Shanghai 200093,China)

机构地区：[1]上海理工大学机械工程学院,上海200093

出　　处：《软件导刊》2022年第10期149-155,共7页Software Guide

基　　金：国家自然科学基金项目(51777126)。

摘　　要：为解决大规模、存在可变障碍物的复杂变电站环境下的自主导航问题,设计一个变电站运动规划框架,包括建立变电站地图、动态避障与局部路径规划、可变障碍物比对及判断,并提出一种基于层次搜索空间尺度的HHA*路径规划方法,能够有效地生成平滑、安全的路径。与现有方法相比:①提出一种新的HHA*方法,首先将变电站空间划分为一个粗尺度,建立常规路径以生成全局目标,然后在精细尺度搜索空间中采用新的启发式函数,以提高生成路径的平滑性;②提出一种新的障碍物称为可变障碍物。在各种复杂的模拟场景和具有挑战性的现实世界中验证该方法的安全性、平滑性、灵活性和高效性,结果表明,该方法的计算效率相比混合A*算法提高了1.83倍,在全局规划阶段相比混合A*算法提高了45.48%,在局部规划阶段相比QBC算法提高了22.87%,相比TEB算法提高了16.37%。证明采用该方法在大规模变电站复杂场景中进行巡检,可满足路径规划平滑性和安全性的要求。In order to solve the problem of autonomous navigation in large-scale complex substation environment with variable obstacles,a substation motion planning framework is designed,including the establishment of substation map,dynamic obstacle avoidance and local path planning,variable obstacles comparison and judgment.In this study,an HHA*path planning method based on hierarchical search spa⁃tial scale is proposed,which can effectively generate smooth and safe paths.The novelty of this work compared to existing methods is twofold:①Proposing a new HHA*method consists of two phases.First,the substation space is divided into a coarse scale,and a conventional path is established to generate global goals.Then,a new heuristic function is used in the fine-scale search space to improve the smoothness of the gen⁃erated path;②A new kind of obstacle called variable obstacle is proposed.Variable obstacle refers to passing through it from different direc⁃tions and can present the characteristics of obstacle or non-obstacle.We verify the security,smoothness,flexibility and efficiency of the pro⁃posed method in a variety of complex simulation scenarios and challenging real-world scenarios.The results show that the computational effi⁃ciency is 1.83 times higher than that of the hybrid A*algorithm.In the global planning stage,it was 45.48%higher than mixed A*,22.87%higher than QBC and 16.37%higher than TEB in the local stage.This method can effectively carry out inspection in large scale substation com⁃plex scene to meet the smoothness and security of the path planning.

关 键 词：自主导航 巡检机器人 路径规划 障碍物 HHA* 变电站 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置]