Non-stationary And Non-gaussian Stochastic Process Simulation And Its Application In Seismic Reliability Of Structures

The safety of engineering structure is the core problem in the field of civil engineering,which is directly related to the safety of people's life and property.As a result,structural reliability theory plays an important role in structural design and safety assessment.Engineering structures inevitably receive random dynamic action,such as strong winds,earthquakes,waves in their useful life.Howerver,those random loads owns non-stationary or ono-Gaussian characteristics,and the existing random dynamic load simulation pay more attention on independent thinking of non-stationary and non-gaussian.For non-stationary non-Gaussian random process,because it not only need to consider the statistical properties of time-varying,but the probability density function of normality,the corresponding simulation research is relatively less.Therefore,based on the current situation,this paper first studies on non-stationary and non-gaussian stochastic process simulation methods,in an attempt to find an efficient and applicable simulation method.In addition,the dynamic system reliability problem of complex structure is also a big problem in the field of reliability research,which involves complex dynamic analysis,failure mode identification and comprehensive failure probability calculation.Therefor,a high-efficient,high-accuracy and concise method for dynamic system reliability analysis is an urgent demand to solve the problem of the similar projects.This paper mainly focuses on the study of non-stationary non-gaussian stochastic dynamic load simulation and dynamic system reliability,proposing a new non-stationary non-gaussian random process simulation method,studying on dynamic system reliability based on the combination of equivalent extreme value event and the probability density evolution method.Finally,it is extended to study of the seismic reliability of transmission towers.The main works and results of this thesis are summarized as follow:(1)Using numerical method based on Mehler's foumula to solve the equivalent correlation coefficient,the peper simulated non-stationary non-Gaussian stochastic process.In the common non-gaussian stochastic process simulation based on memoryless nonlinear transformation,the core work is the transformation from the potential gaussian stochastic process to the fitting non-gaussian stochastic process,and the most important thing is the transformation of the correlation coefficient between them.The traditional calculation method of correlation coefficient transformation involves the solution of two-dimensional integral problem,which means a large calculation and limited application scope.In this paper,Mehler's formula is first introduced to solve the conversion problem of equivalent correlation coefficient in non-gaussian simulation.Then combined with random function-spectral representation method of random function,the simulation of non-gaussian process is realized efficiently and accurately.On the basis of the classical model of non-stationary ground motion,non-gaussian characteristics are introduced by means of memoryless nonlinear translation method,and the non-stationary ground motion excitation with non-gaussian characteristics is simulated.Finally,the feasibility and accuracy of the proposed method are verified by the classical seismic power spectrum.(2)Probability density evolution method(PDEM)through introducing the asymptotic delta sequences of the Dirac's function is used to solve the problem of structural dynamic system reliability under non-stationary and non-gaussian ground motions.When it comes to classic dynamic reliability analysis,it is difficult to avoid introducing kinds of cross assumptions,such as Poisson hypothesis,Markov hypothesis,and so on,resulting in poor convergence or small scope of application,etc.In addition,the reliability solution of complex structures involves the identification of multiple failure modes,which makes it very difficult to calculate the reliability of dynamic systems.In this paper,on the basis of the classical dynamic reliability analysis theory,the equivalent extreme event theory is introduced,and PDEM based on delta sequence method is proposed to solve the dynamic system reliability efficiently and accurately.Finally,by analyzing the comparison of random seismic response and seismic reliability between seven layers single span steel frame structure under non-stationary seismic load and non-stationary non-Gaussian seismic load,the accuracy of proposed method are verified and the rationality of introducing non-Gaussian charateristics is verified from the perspective of reliability.(3)Random seismic response and seismic reliability analysis of UHV transmission towers under non-stationary non-gaussian ground motion loads.At present,there are many analyses on UHV transmission tower structure under strong wind load.In this paper,the random seismic response of HVDC transmission tower in northwest Yunnan province under non-stationary non-gaussian ground motion load is realized by ANSYS finite element software.Based on the proposed method in(2),the dynamic reliability of transmission tower structure system is calculated,which provides a new idea for seismic reliability analysis of transmission tower structure.