Abstract: The intelligence level of daily inspection tasks might be greatly improved by using robots to measure the temperature of high-voltage lines in substations, but the problems of road network limitation and energy consumed by frequent start-stops were encountered in robot inspection processes. And the mutual restriction between the optimal path and the specific temperature measurement location might be considered. In order to solve the above problems, the substation environment was modeled based on graph theory, and a path planning model with the goal of minimizing time consumption was established by analyzing the pose constraints of the camera when the robot measured the temperature of the target in the road network. Then, a path optimization method was proposed based on the improved ant colony algorithm and the method of selecting inspection stops was also presented based on greedy clustering. And the optimal inspection paths of the robots, the positions of the stops and the sets of temperature measurement points corresponding to each stops were iteratively obtained through ant colony algorithm. Finally, simulation cases were used to verify the feasibility and robustness of the algorithm, and the experimental results of the inspection robots show that the method may be used for substation inspection.

Key words:  , inspection robot, path planning, ant colony algorithm, clustering algorithm

摘要： 采用机器人对变电站高压线路测温可以大大提高日常巡检任务的智能化水平，但是巡检过程中存在路网限制和频繁启停能量消耗大等问题，须考虑最优路径与最佳测温停靠地点的相互制约性。为了解决上述问题，基于图论对变电站环境进行建模，通过分析巡检机器人在路网中对目标点测温时相机的位姿约束，建立了以时间消耗最小为目标的路径规划模型。提出了一种基于改进蚁群算法的路径优化方法和基于贪婪思想聚类的巡检停靠点选取方法，通过蚁群算法迭代获得机器人的最优巡检路径、巡检停靠点序列以及各停靠点对应的测温点集。最后，采用仿真案例验证了算法的可行性和鲁棒性，基于巡检机器人的实验结果表明该路径规划模型可用于变电站巡检。

关键词: 巡检机器人, 路径规划, 蚁群算法, 聚类算法