Reliability Analysis Of Operators Of A Digital Nuclear Power Plant Based On Semantic Analysis Model

With the continuous development of information technology and computer technology,more and more nuclear power plant main control rooms have adopted advanced digital control systems(DCS).In a digital nuclear power plant,the operator's monitoring of the power plant status and control of the power plant system is mainly done on the computer screen to replace the traditional control room in the disk table,instrument panel,signal lights,buttons,switches and other equipment.The digitization of the human-machine interface will cause changes in the cognitive behavior of the operator,resulting in new human errors and affecting the safe operation of the nuclear power plant.Traditional Human Reliability Analysis(HRA)calculates the reliability of a person by analyzing the probability of the operator's behavioral action(i.e.,“doing”)success or failure.The digitization of nuclear power plants mainly affects the cognitive behavior process of operators.Therefore,the argument of this paper starts from the perspective of the operator's cognitive behavior(i.e.,“thinking”)and studies the reliability of theoperator.Semantics is a kind of psychological representation of cognitive behaviors,an internal understanding of objective things,and also a reflection of external knowledge information in individual psychology.This paper points out the importance of the operator's semantic understanding by analyzing the human-machine interface characteristics of the digital control room,the cognitive behavior characteristics of the operators and the cognitive semantic features.Then the cognitive and semantic framework of the operator of the nuclear power plant is proposed,and then the semantic analysis process is described,and the semantic analysis model of the operator is established.The model was applied to operator training,and the functional effects of operator semantic training were verified by DCS simulator experiments.The results show that under the accident conditions of nuclear power plants,the training based on the semantic qualitative analysis model can make the operator more clearly understand the knowledge structure of the human-machine interface of DCS and help to consolidate the mental model and the situation model,and perceive the information of the operator.A series of cognitive behaviors such as processing and processing have a more positive impact.In addition,using the mathematical method of Bayesian network,the process of semantic quantitative analysis model deductive operator error diagnosis isestablished.The example verification shows that the semantic analysis model can also provide new evaluation for the quantitative evaluation of cognitive reliability of operator diagnosis decision.Ideas to improve the safety and reliability of nuclear power plant operations.