Research Requirements of Active Discovery and Abduction in Distribution Network Maintenance Risk

Identifying risks in distribution networks is of great significance for preventing and discovering faults, and development of maintenance plans. The current challenge is how to decrease recognition errors and improve the accuracy and efficiency of recognition algorithms when processing massive amounts of data, especially sample data is limited or imbalanced. Therefore, this paper tries to investigate human-machine interaction and human experience to compensate for the uneven quantity and distribution of risk identification data in the process of maintenance plan formulation and pattern plan generation. This paper analyzes the current status and progress of the research and also gives the technical requirement framework.