Research On The Evaluation Of Human-machine-interface In Nuclear Power Plant Based On The Human Reliability Analysis

The main control room of the nuclear power plant is the main place for the operator to monitor and control the safe production of the nuclear power plant,and the human-machine interface is the carrier of the man-machine interaction between the operator and the nuclear power plant system.For the complex and high-risk systems such as nuclear power plants,there has been a consensus on the importance of improving safety and operator's performance,and the poor human-machine interface has become one of the main causes of human failures or accidents.Advanced nuclear power plant control room based on digital technology have advanced instrumentation(I&C)system,computerized human-machine interface compared to the traditional human-machine interface,due to the improvement of the degree of automation,there have some considerable changes and new features,such as computerized alarm systems,information display systems,computerized procedure systems,soft control,and interface management etc.Those new features may have some negative impact on human error.These changes have brought new challenges to the human-machine interface of the advanced control room,and also become an important research topic on how to identify the potential human-machine interface defects which can easily lead to human error or weaken the performance of operators.The goal of this paper is to evaluate the existing human-machine interface by human reliability analysis(HRA)method,to identify the weakness and defects of human-machine interface,and to provide decision support for human-machine interface modification and optimization.The main research contents are as follows:1.Research the influence of the development of computerized human-machine interface on the HRA method.Analyze the differences and characteristics of the computerized human-machine interface between the traditional human-machine interface.Mainly discuss and analysis the reasons of optimizing the human-machine interface in the main control room of nuclear power plant.2.Discuss the human factor of the operators according the characteristics of the human-machine interface in the main control room of the nuclear power plant.3.Study the existing HRA methods,and focus on the CREAM method as an example to illustrate the process of CREAM method to calculate the human error probability.Analyze the problems and shortcomings of CREAM method combined with the characteristics of computerized human-machine interface of nuclear power plant.4.Considering the advantages of Bayesian network in dealing with the relationship between common performance conditions(CPC),this paper presents an improved CREAM method to guide human error probabilistic quantification calculation based on fuzzy logic + Bayesian neural network.5.Taking the operation scene of one nuclear power plant as an example,discuss the probability of human error when the operator handles the Mistake Safety Injection scene through the human-machine interface.The improved CREAM method is used to calculate the probability of human error and the validity of improved CREAM method has been verified.6.Analyzes the feasibility of checking the main defects of the human-machine interface involved in the error-safety injection scenario based on the human factor engineering check list table(HEC),and gives some suggestions and measures to improve the reliability of the human-machine interface of the computerized control room.