Study on Intelligent Diagnosis Algorithm Based on the Diagnosis Model of Accidents in Digital NPPs

Authors

  • Lingyan Gu, Yichen Zhang, Yuhang Li, Yanqi Liu

Abstract

Developing intelligent diagnosis system and human performance support system is significant in human-related researches in nuclear power plants (NPPs). The main causes of safety accidents in NPPs not only include the single or superimposed occurrence of various conventional accidents but the misjudgment from operators. In this paper, the intelligent diagnosis algorithm based on the previous study of the intelligent diagnosis model of accidents constructed by event-oriented procedure (EOP) in digital NPPs is further studied and has already been simulated by Monte Carlo simulation, which was realized by establishing the accident-state correlation matrix and the accident-state probability matrix, then matching the monitored result vector with each accident feature vector. In order to find the optimal solution in case that there are more than two similar distances between the monitored result vector and several accident feature vectors simultaneously, the virus algorithm is presented necessarily in this paper. Finally, the simulation results represent that the correct diagnosis coverage rate reaches 100%, which also mean the algorithm can provide proposal with diagnosis decision for operators in NPPs.

Published

2020-03-31

Issue

Section

Articles