Research And Application Of Deformation Analysis Model For Nuclear Power Plant

China's rapid development of electricity,with the continuous innovation of science and technology and people increasingly aware of resource conservation,environmental health,the importance of circular economy,nuclear power because of its large capacity,small footprint,stable power generation,pollution-free advantages Just to fill China's power gap.Most of the nuclear power plant engineering buildings are built on the bedrock,but in the construction,after completion and operation by the impact of various loads,as well as the power station aging and other reasons is bound to deformation,deformation or rate to a certain extent,will affect The normal operation of the nuclear power plant that endanger nuclear safety and major safety incidents.Therefore,it is necessary to analyze and study the deformation monitoring data in time,and to predict the trend of deformation trend accurately so that the corresponding measures can be taken in time to avoid the accident.The nuclear power plant is currently no systematic solution to the management of deformation monitoring.Therefore,the monitoring data is managed by a set of deformation monitoring and management system,and a forecasting method is used to forecast the data.In this paper,the long-term data collection and data collection and finishing work of nuclear power plants,on the basis of in-depth understanding of nuclear power plant deformation monitoring,access to relevant information,study and study of deformation monitoring and forecasting theory and methods,and the current major forecasting model This paper introduces the methods and theories of commonly used prediction model,introduces the principles and characteristics of regression analysis,gray model theory and neural network model,and analyzes the modeling mechanism of the combined model and designs and prepares the deformation monitoring and management system.Mainly include the following aspects:(1)analysis and research data analysis and forecasting models and methods,including regression analysis,gray model,neural network model,and analysis of its characteristics and accuracy.(2)Introduce the principle of combinatorial model and put forward the GM-Logistic model,and analyze and study the two methods of right and variable right.(3)Introduce the design and realization of the deformation monitoring and management system of the nuclear power plant and the function of each module,and analyze the deformation and settlement data of the conventional island plant to verify the reliability of the system.Finally,the results of GM(1,1)model,Logistic model and GM-Logistic fixed combination and variable weight combination model are compared.The prediction accuracy and effect of the combined model are higher than that of single model and variable weight combination method The combination is more reasonable.