Research On Key Technologies Of Dynamic Reliability Monitoring In Nuclear Power Plant

Monitoring the status of systems and equipment is a routine means of ensuring the safe operation of nuclear power plants.In recent years,the online risk monitoring technology represented by the probabilistic safety assessment(Living PSA,LPSA)based on the reliability model update technology has begun to be applied in the operation of nuclear power plants.This strategy not only focuses on changes in system and equipment states,but also assesses the risk of these state changes to the reactor core damage,so that plant operational risks remain as knowable and controllable as possible.The application of the existing nuclear power plant online risk monitoring technology is mainly aimed at the power plant process system.Process considerations are relatively rough.At the same time,the existing online risk monitoring technology of nuclear power plants only analyzes the state changes of systems and equipment that have a significant impact on the operational safety of nuclear power plants,and does not pay enough attention to non-safety power plant systems and equipment.In addition,with the localization of nuclear power equipment and the development and application of automatic control theory in nuclear power plant instrumentation and control technology,the human-computer interaction process in nuclear power plants has become more complex.In order to ensure the economy and safety of the power plant,it has become a new research hotspot to pay attention to the reliability monitoring technology of the operation of the nuclear power plant.The technical means of reliability monitoring and risk monitoring are interoperable.In order to solve the above technical problems,this paper has carried out the research on the key technology of dynamic reliability monitoring of nuclear power plants.First,by analyzing the interaction process between the various elements of the power plant,a scheme for dynamic monitoring of the reliability of nuclear power plants is proposed,and the modeling scope is further expanded from the process system of the power plant to the instrument control system,power plant personnel and their mutual interaction.The coupling relationship makes the reliability monitoring more relevant to the actual operation of the power plant and reflects the process reliability results of the power plant.Secondly,in view of the difficulty of equipment reliability data evaluation in the process of localization of nuclear power equipment,a complex equipment reliability data update algorithm is given.On the basis of the degradation path analysis(PACE)method,a degradation path functional based The reliability prediction algorithm of analysis and similarity fitting is suitable for reliability data prediction of small sample equipment in nuclear power plants.Thirdly,the modeling and analysis method of GO-FLOW is improved,and a new GO-FLOW operator is proposed and expanded,which is used to express the common reliability logic of automatic control process,and the reliability of the new operator is given.Mathematical expression.Through the improved GO-FLOW model,the coupled reliability modeling and analysis of the automatic control process of the instrumentation and control system and the functional structure of the process system are realized.In addition,in view of the interaction process between power plant personnel and process system and instrument control system,the analysis and research on the operator’s cognitive behavior process are carried out.The process of information acquisition,information diagnosis and information reasoning completed by the operator in the process of cognitive behavior is analyzed,an evidence inference model based on Bayesian network is proposed,and the evidence reasoning algorithm in the cognitive behavior process of operators under abnormal working conditions of power plant is designed.Finally,aiming at the hysteresis of model update in LPSA technology,research on dynamic reliability monitoring method based on evidence update technology is carried out.The power plant operation information monitored by the I&C system is designed as direct evidence,the information confirmed in the process of operator’s cognitive behavior,and the information obtained from the evidence inference model analysis are used as indirect evidence to jointly update the reliability monitoring model.The research on the update algorithm of the GO-FLOW model of the power plant system is carried out and the common cause problem of indirect evidence is dealt with.The simulation verifies the dynamic reliability monitoring process of the power plant under the real-time cognitive state of the operator.The research results of this paper provide new ideas and solutions for solving the key problems in the dynamic reliability monitoring of nuclear power plants,and enrich the theory and methods of dynamic reliability monitoring of nuclear power plants.Key technologies provide theoretical support for reliability data evaluation and update methods of complex equipment in process systems;provide technical support for solving coupled reliability modeling between power plant components;analysis and evidence for operator cognitive behavioral processes The realization of inference provides a feasible solution for modeling and algorithm;it provides technical support for the updating technology of reliability monitoring model in line with the operator’s real-time cognitive state.