Reliability Analysis Method Of Concrete Gravity Dam System Based On Structural Performance Feedback

The load and material parameters of concrete gravity dams during the design reference period are uncertain.Uncertainty includes randomness,fuzziness,and cognitive uncertainty,and the impact of various types of uncertainty on concrete gravity dams is also different.Therefore,carrying out the impact of multi-source uncertainty on the service safety of concrete gravity dams will provide strong support for the reinforcement of dangerous dams and the design and construction of high dams in China.Based on the design data and prototype monitoring data of a concrete gravity dam and considering the impact of multi-source uncertainties on the service safety of gravity dam,this paper updates the distribution of uncertain parameters by carrying out probabilistic inversion and interval inversion of mechanical parameters of gravity dam based on prototype monitoring data,using stochastic mathematics,interval mathematics,surrogate model,optimization algorithm,Monte Carlo method and active learning strategy,A systematic study has been conducted and efficient calculation methods for the reliability indicators of gravity dams,failure mode search methods,and system reliability calculation methods have been proposed,providing a new approach for reasonably evaluating the safety and stability of gravity dams during their service life.The main content is as follows:(1)In view of the difficulties in prototype and laboratory tests and poor representativeness of sample data when obtaining parameters,and considering the functional relationship between prototype monitoring data and some physical and mechanical parameters of gravity dam,a surrogate model of deformed hydraulic pressure component of concrete gravity dam is constructed based on the moving least square method optimized by differential evolution algorithm,A probability inversion method for mechanical parameters of gravity dams based on experimental data and an interval inversion method for mechanical parameters of gravity dams considering measurement errors have been proposed,achieving efficient updating of uncertain parameters.(2)On the premise of drawing up the failure mode of gravity dam by referring to engineering experience and relevant specifications,and taking full account of the impact of multi-source uncertainties on the service safety behavior of concrete gravity dam,based on the information entropy method and active learning Kriging model,a hybrid reliability analysis method of probability fuzzy interval for gravity dam is proposed to efficiently calculate the failure probability interval of the proposed failure mode,And combined with the general boundary method to calculate the failure probability of the gravity dam system,it provides a way to efficiently diagnose the service safety of concrete gravity dams.(3)Considering that the concrete material parameters of gravity dam in complex environment are discrete and the failure situation is very complex,the proposed failure mode is not scientific and feasible enough.At the same time,in order to further improve the calculation efficiency of failure probability,on the premise of clarifying the failure criteria of gravity dam unit,a calculation method of reliability index of gravity dam unit is proposed,which integrates the information entropy method,KKT optimization conditions and JC method,Improving the branch and bound method applicable to frame structures,searching for the main failure modes of gravity dams,and calculating the reliability of gravity dam systems using the narrow bound method,provides a method for scientifically diagnosing the service safety of concrete gravity dams.