基于图论的电力巡检机器人智能寻迹方案

An Intelligent Patrolling Robot Addressing Program Based on Graph Method

针对无人值守变电站遥视系统的网络化和智能化多点监测趋势,结合巡检机器人技术,提出一种基于图论的智能寻迹方案。将实际检测现场以图的形式转化为电子地图信息,建立节点关联矩阵,利用基于传递闭包理论的路径搜索算法对各检测点关联信息进行关联路径搜索。通过现场运行,对实际工作中遇到的问题进行调试,并针对这些问题对方案进行改进。该方案具有避免搜索死循环、占用资源少、计算速度快的优点,适合现场检测的实时性要求。通过应用于多传感检测的遥视系统,证明了该方案的高效性和可行性。

An intelligent addressing program based on the graph theory is adopted in the patrolling robot technology to deal with the multiple intelligent network trend in the monitoring system in unattended substations. The geographical information is represented as an electronic map with nodes correlation matrix. The path searching algorithm based on the transfer closure theory is used to search for the correlation information at all the measuring points. Problems encountered in operation are tested and resolved, with the scheme improved to avoid similar problems in future. With such advantages as freedom from endless loop search, little demand for resources, and high computing speed, this scheme is suitable for on-the-spot real-time test. The high efficiency and feasibility of the scheme is verified through its application in a remote television system using multi-sensor detection.