基于机器视觉和进化算法的变电站巡检机器人运行路径规划方法  

Operation Path Planning Method of Substation Inspection Robot Based on Machine Vision and Evolutionary Algorithm

作　　者：朱大智 颜清 王柳乃 ZHU Dazhi;YAN Qing;WANG Liunai(Information and Communication Branch of Hainan Power Grid Co.,Ltd.,Haikou 570100,China)

机构地区：[1]海南电网有限责任公司信息通信分公司,海南海口570100

出　　处：《微型电脑应用》2024年第12期202-206,共5页Microcomputer Applications

摘　　要：变电站巡检机器人在路径规划的过程中,由于路径障碍处理不当,导致规划精度以及适应性较弱,整体规划效果较差。针对该问题,基于机器视觉和进化算法构建巡检环境模型,加强对内部传感系统的管理,分析在不同情况下的机器视觉传导状况,检测与处理障碍图像数据,提升初始数据处理的精准程度和路径规划的效率,进而增强规划的鲁棒性,提升规划精度。实验结果表明,所提方法相较于其他2种方法在复杂的环境下具有更强的能控性和鲁棒性,机器人运动轨迹无碰撞,适应度值整体保持在21以上,能够有效实现变电站巡检机器人运行路径规划。In the process of path planning,due to improper handling of path obstacles,the substation inspection robot has weaker planning accuracy and adaptability,and the overall planning effect is poorer.Aimed at this problem,an inspection environment model is constructed based on machine vision and evolutionary algorithm.The management of the internal sensing system is strengthened,the machine vision conduction status under different circumstances is analyzed,the obstacle image data is detected and processed,the accuracy of initial data processing and the efficiency of path planning are improved.Hence,the robustness of planning and the accuracy of planning is enhanced.The experimental results show that the proposed method has stronger controllability and robustness than the other two methods in complex situations,the robot motion trajectory is collision-free,the overall fitness value remains above 21,which can effectively realize the operation path planning of substation inspection robot.

关 键 词：机器视觉 进化算法 变电站巡检机器人 路径规划 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]