Study On Safety Evaluation Method Of Gravity Dam Based On Game Theory And Improved Cloud Model

Gravity dams occupy a very important position in water conservancy projects,and while bringing enormous social and economic benefits to China’s development,there are also risks and hidden dangers that cannot be ignored.Once the damages happen to the gravity dams,they may lead to flooding,loss of water resources,causing significant safety losses of life and property,and affecting regional stability.The safety monitoring of gravity dams is a key segment in understanding their safety behavior.As the main content of reservoir safety management and an important measure for risk control,it provides a data basis for safety assessment.Therefore,the research on gravity dam safety assessment methods based on monitoring data should receive higher attention.The purpose of this article is to construct a gravity dam safety evaluation model based on game theory and improved cloud model by integrating the comprehensive weighting method of game theory with the improved cloud model based on MBCT-SR algorithm,and to carry out research on gravity dam safety evaluation methods based on monitoring data,in combination with multi-year monitoring data of reservoir gravity dams.The main research contents of this article are as follows:(1)Starting with the structure and construction technology of concrete gravity dams,the characteristics of concrete gravity dams are analyzed;Study the failure mechanism of gravity dams,analyze various factors that affect the safety performance of concrete gravity dams,as well as the possible impact of extreme environmental and effect quantities on the safety of gravity dams,and provide a basis for selecting indicators and constructing a safety evaluation system for gravity dam safety evaluation based on monitoring data.(2)Based on the operation of the monitoring system for the gravity dam of the Longgang Reservoir,and the quality of existing monitoring sequences,the possible problems with monitoring data are analyzed,and a reasonable research timing is selected accordingly;SPSS software was used to verify the normality of monitoring data based on nonparametric testing methods and image testing methods;Using the Laida Criterion(3σ Criteria)Identify and eliminate singular values from the original monitoring data within the selected time series,and analyze those which were resulted from actual physical factors;Based on the spatiotemporal correlation of the gravity dam itself,a regression statistical model is used to predict and interpolate the missing monitoring data at each measurement point,and the feasibility of this method is verified;Using MATLAB software,based on appropriate wavelet basis functions,select a reasonable decomposition level,and perform signal-noise separation processing on the data,eliminating errors caused by noise effects,thereby obtaining a sequence which is more realistic.(3)Based on the MBCT-SR algorithm,the random sampling method is used to extract digital features from the sample data,and then calculate the cloud parameters of each measurement point data.Through comparison with the traditional single-step reverse cloud transformation algorithm,the advantages of the multi-step reverse cloud transformation algorithm are proven,which means the former algorithm has better ability to achieve quantitative and qualitative evaluation conversion.Taking the apparent horizontal displacement monitoring project of the gravity dam of Longgang Reservoir as an example,the monitoring data is regarded as cloud droplets at each measuring point,the cosine similarity method is used to calculate the cloud similarity between different measurement points,and the cloud similarity coefficient is obtained through normalization processing.Based on this,the correlation of displacement monitoring data from multiple measurement points is analyzed,and the measurement point with the highest correlation degree with the overall deformation behavior is identified.(4)Combined with the actual situation of the gravity dam monitoring system of the Longgang Reservoir in Banan District,a multi-level safety evaluation system for gravity dams based on monitoring data is constructed.Based on game theory,subjective weight and objective weight are integrated,and comprehensive weight coefficients are assigned to evaluation indicators at all levels.Based on the improved cloud model,the normal cloud safety evaluation criteria for each level of evaluation indicators are constructed,and then a safety evaluation model for the gravity dam of Longgang Reservoir based on game theory and improved cloud model is constructed;The relative closeness calculation method is used to determine the risk level based on the relative closeness of the actual monitoring data and each rating according to the threshold range of the evaluation indicators at each level of the safety evaluation system.The safety characteristics of the environmental quantity,displacement effect quantity,and seepage effect quantity during the dam operation period are obtained,providing a model drive for the gravity dam safety evaluation based on the monitoring data,and realizing the conversion between quantitative and qualitative evaluation.