基于双目视觉的电力巡检机器人轨迹跟踪方法研究  

Research on trajectory tracking method of power inspection robot based on binocular vision

作　　者：佟芳 马晓琴 李文明 Tong Fang;Ma Xiaoqin;Li Wenming(State Grid Qinghai Information&Telecommunication Company,Qinghai Xining,810000,China)

机构地区：[1]国网青海省电力公司信息通信公司,青海西宁810000

出　　处：《机械设计与制造工程》2025年第1期40-44,共5页Machine Design and Manufacturing Engineering

摘　　要：电力巡检机器人轨迹跟踪在线测距中存在延时与可靠性较差等问题,导致轨迹跟踪精度较低。为此提出基于双目视觉的电力巡检机器人轨迹跟踪方法。首先,构建机器人运动学模型分析运动误差,获取点云数据并处理点云网格信息,加强对轨迹数据的收集力度。然后,提取相应运动轨迹路径,按照相关路径检验机器人轨迹跟踪方法的可行性,降低轨迹跟踪信息的提取失误率。最后调整轨迹跟踪系统角度,建立双向导航系统以实时取得跟踪结果。经实验验证,所提方法轨迹跟踪精度较高,跟踪误差在2%以下,能够有效地完成电力巡检机器人轨迹跟踪。There are problems such as delay and poor reliability in the online distance measurement of power inspection robot trajectory tracking,resulting in low trajectory tracking accuracy.A trajectory tracking method for power inspection robots based on binocular vision is proposed for this purpose.Firstly,it constructs a kinematic model of the robot to analyze motion errors,obtains point cloud data and process point cloud mesh information,and strengthens the collection of trajectory data.Then,it extracts the corresponding motion trajectory path and verifies the feasibility of the robot trajectory tracking scheme according to the relevant path,reduces the error rate of extracting trajectory tracking information.Finally,the angle of trajectory tracking system is adjusted,and a two-way navigation system is established to obtain real-time tracking results.Through experimental verification,the proposed method has high trajectory tracking accuracy,with a tracking error of less than 2%,and can effectively complete the trajectory tracking of power inspection robots.

关 键 词：双目视觉技术 电力巡检 巡检机器人 轨迹跟踪 机器人轨迹 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]