Abstract:The inspection of distribution network room is an important means to comprehensively find out the hidden dangers of equipment defects, provide the basis for test and inspection, and improve the operation level of distribution network. In view of the prominent problems of the wide range of distribution website rooms, large inspection workload and the lack of front-line operation and maintenance personnel, an inspection strategy of distribution website rooms considering temporal and spatial distribution is proposed. In this method, the coordinates of the station buildings within the operation and maintenance area are derived from the geographic information system (GIS) and preprocessed to form the data source file. Considering the characteristics of the actual application scenarios such as the carrying capacity of the operation and maintenance personnel, the importance and intelligence of the station buildings, and the length of the operation and maintenance cycle, K-means algorithm and genetic algorithm are used to cluster and optimize the preprocessed data, forming the station building patrol strategy under the regular patrol mode and the station building differential patrol strategy considering the length of the cycle. The example analysis shows that the proposed method can effectively optimize the inspection path, guide the operation and maintenance personnel to carry out the inspection work, and improve the inspection quality and efficiency.