A Sensor Fault-Tolerant Accident Diagnosis System.

Jeonghun Choi,Seung Jun Lee

doi:10.3390/s20205839

Abstract

Emergency situations in nuclear power plants are accompanied by an automatic reactor shutdown, which gives a big task burden to the plant operators under highly stressful conditions. Diagnosis of the occurred accident is an essential sequence for optimum mitigations; however, it is also a critical source of error because the results of accident identification determine the task flow connected to all subsequent tasks. To support accident identification in nuclear power plants, recurrent neural network (RNN)-based approaches have recently shown outstanding performances. Despite the achievements though, the robustness of RNN models is not promising because wrong inputs have been shown to degrade the performance of RNNs to a greater extent than other methods in some applications. In this research, an accident diagnosis system that is tolerant to sensor faults is developed based on an existing RNN model and tested with anticipated sensor errors. To find the optimum strategy to mitigate sensor error, Missforest, selected from among various imputation methods, and gated recurrent unit with decay (GRUD), developed for multivariate time series imputation based on the RNN model, are compared to examine the extent that they recover the diagnosis accuracies within a given threshold.

Highlights

In safety-critical systems, a prompt reaction to anomalies is a crucial factor to minimize any related consequences
A wrong diagnosis would lead to selecting the wrong optimal recovery procedures (ORPs), which could result in multiple human errors
In the nuclear field, where safety is of utmost importance, even though nuclear power plants (NPPs) are equipped with numerous autonomous safety systems, responses to abnormal states still largely depend on the judgments of plant operators

Summary

Introduction

In safety-critical systems, a prompt reaction to anomalies is a crucial factor to minimize any related consequences. Diagnosis is crucial because it determines the particular optimal recovery procedures (ORPs) that contain the essential mitigation tasks [1,2]. Diagnosis procedures give intuitive logics for identifying accidents based on a series of symptom checks, but can be a demanding task for the plant operators because the early phases of an emergency situation may affect the accident consequences. In this regard, a wrong diagnosis would lead to selecting the wrong ORPs, which could result in multiple human errors

Results

Discussion

Conclusion