500kV超高压变电站安全巡视路径自动化寻优研究

Study on Automated Optimization of Safety Patrol Paths in 500 kV Ultra-High Voltage Substations

变电站安全巡视路径的拐点较多、定位技术存在差异、避障效果较差,直接造成了最优路径长度过长的问题。对此,提出基于机器视觉的500 kV超高压变电站安全巡视路径自动化寻优方法。通过背景差法检测500 kV超高压变电站的场景信息,构建安全巡视环境模型。以机器视觉原理为基础,创新性地结合Hessian矩阵设计巡视机器人定位技术。根据安全巡视机器人当前定位信息,采用机器视觉和人机交互的模式,获取初始变电站安全巡视路径。建立以最短路径长度为目标的寻优目标函数,利用粒子群优化(PSO)算法自动求解出最优巡视路径。仿真试验结果表明:搜索最优路径得以缩短。该研究保证了安全巡视过程的完整性和智能化。

The substation safety patrol path has more inflection points,differences in localization technology,and poor obstacle avoidance effect,which directly causes the problem of the optimal path length is too long.In this regard,an automated optimization method of 500 kV ultra-high voltage substation safety patrol path based on machine vision is proposed.The scene information of 500 kV ultra-high substation is detected by background difference method,and the safety inspection environment model is constructed.Based on the principle of machine vision,the inspection robot localization technology is innovatively designed by combining with Hessian matrix.According to the current localization information of the safety inspection robot,the machine vision and human-machine interaction modes are adopted to obtain the initial substation safety inspection path.The optimization objective function is established with the shortest path length as the objective,and the optimal patrol path is obtained by using the particle swarm optimization(PSO)algorithm.Simulation text results show that the optimal path for searching is shortened.The study ensures the integrity and intelligence of the safety patrol process.