A Personalized Alarm Distribution Approach Based on Collaborative Filtering in New-type Power System

Abstract

Due to the complexity and heterogeneous characteristics of the information system architecture of the new-type power system, once a fault occurs during operation, a large number of alarm information will be generated. In order to improve the efficiency of operation and maintenance (O&M) personnel and minimize the loss caused by failures, a collaborative filtering algorithm is introduced to provide personalized alarm information distribution. Each alarm is preferentially assigned to O&M personnel with high processing probability, so that O&M personnel can receive alarm information they can handle. In addition, the minimum distribution number is limited to ensure that each alarm can be received by enough O&M personnel. Finally, experiments show that the proposed method can effectively improve the O&M efficiency and improve the fault recovery capability of the new-type power system.