变电站巡检机器人视觉导航路径快速提取方法  

Fast extraction method of visual navigation path for substation inspection robot

作　　者：刘兴华 LIU Xinghua(Zibo Power Supply Company of State Grid Shandong Electric Power Company,Zibo 255000,China)

机构地区：[1]国网山东省电力公司淄博供电公司,山东淄博255000

出　　处：《电子设计工程》2024年第13期23-26,31,共5页Electronic Design Engineering

摘　　要：针对变电站机器人实际巡检节点与预设节点间出现明显偏离的问题,为了在生成导航路径时使控制主机能够快速提取节点对象,提出变电站巡检机器人视觉导航路径快速提取方法。通过横轴线、纵轴线、空间轴线设置变电站巡检机器人视觉坐标系,提取离散特征点,并对提取的离散特征点进行拟合,完成变电站巡检机器人视觉导航路径检测;根据检测结果,求解巡检导航参数,并以此为依据建立路径节点运动模型,确定模糊提取系数,实现变电站巡检机器人视觉导航路径快速提取方法的设计。对比实验结果表明,快速提取方法将实际巡检节点与预设节点之间的偏离度控制在5%以内,控制主机不需消耗过多时间辨别巡检节点,符合快速提取导航路径的需求。Aiming at the problem of obvious deviation between the actual inspection node and the preset node of the substation robot,in order to enable the control host to quickly extract the node object when generating the navigation path,a fast extraction method for the visual navigation path of the substation robot is proposed.Set the visual coordinate system of the substation patrol robot through the horizontal axis,vertical axis and spatial axis,extract discrete feature points,and fit the extracted discrete feature points to complete the visual navigation path detection of the substation patrol robot.According to the detection results,solve the navigation parameters,and based on this,establish the path node motion model,determine the fuzzy extraction coefficient,realize the design of visual navigation path fast extraction method for substation inspection robot.Comparative experimental results:The fast extraction method controls the deviation between the actual patrol node and the preset node within 5%,and the control host does not need to spend too much time to identify the patrol node,which meets the requirements of fast extraction of navigation paths.

关 键 词：变电站 巡检机器人 导航路径 离散特征点 运动模型 模糊系数 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]