Nuclear power plant common accident diagnosis method and nuclear power plant common accident diagnosis system

Abstract

The invention discloses a nuclear power plant common accident diagnosis method and a nuclear power plant common accident diagnosis system. The method comprises the following steps: determining an accident feature sequence set and a monitoring result sequence set; performing weight identification processing on the accident feature sequence set to obtain a corrected accident sequence set; performingweight identification processing on the monitoring result sequence set again to obtain a corrected monitoring sequence set; determining a gray correlation degree sequence set according to the corrected accident sequence set and the corrected monitoring sequence set; and finally determining an accident diagnosis result according to the grey correlation degree sequence set. The method solves the problem of high probability errors of nuclear power plant operators under accident conditions, does not depend on the pre-test probability of source accidents, only needs to obtain the logic relationship between accidents and states, avoids the difficulty in obtaining the pre-test probability of source accidents in a nuclear power plant, and is convenient to operate, accurate in diagnosis and easy to implement.

Description

technical field [0001] The invention relates to the technical field of fault diagnosis, in particular to a method and system for diagnosing common accidents in nuclear power plants. Background technique [0002] Nuclear power plants are an important way to utilize nuclear energy, and ensuring the safe operation of nuclear power plants is a very important and thorny problem that human society needs to face and solve. In recent years, although the reliability of nuclear power plant hardware equipment has been gradually improved, and nuclear power plants have installed a number of automatic protection devices for special safety systems, under accident conditions, it is necessary to ensure that the nuclear power plant transitions to a safe state and the safety of nuclear power plants Operator intervention must be relied upon. Statistics show that accidents caused by human factors are about 60%-90%, which is one of the main factors of nuclear power plant accidents. The place wh...

AI Technical Summary

Problems solved by technology

In addition, the diagnosis of faults in many fields is based on the maximum posterior probability criterion of Bayesian theory. Since the pre-test probability of nuclear power system source accidents is difficult to obtain, the diagnosis based on the maximum posterior probability criterion of Bayesian theory method failure

Images