A CONDITION MONITORING APPROACH BASED ON ISOLATION FOREST ALGORITHM FOR NUCLEAR POWER PLANT

Abstract

The nuclear power plant is a complex, non-linear and highly dynamic system, composed of many devices, functional requirements. A reliable condition monitoring technology is one of the important means to improve its safety. The traditional threshold-based condition monitoring method has many shortcomings in practicality. According to the actual situation, the normal operation data is of large amount, and the abnormal data is little. In this paper, a new condition monitoring method based on isolation forest algorithm is proposed to solve this problem and the simulation proves that the proposed method can quickly and accurately identify the condition of the system.