Research On Structural Reliability Analysis Method Based On Intelligent Optimization Algorithm

With the rapid development of science and technology,the engineering structure that people face is becoming more and more complex,and the complex engineering structure may cause people hard to analyze its safety and collapse.Therefore,it is urgent to study the reliability analysis of mechanical structure.Now,the core of reliability research is to solve the reliability index.HL-RF algorithm has been widely used in the early stage because of its advantages of simple and efficient,less memory.However,when the limit state function is nonlinear or has non convex characteristics,the convergence of HL RF iterative algorithm is hard to be guaranteed.This paper presents a hybrid algorithm based on Augmented Lagrangian multiplier and intelligent algorithm by studying traditional reliability optimization method and intelligent optimization method.By introducing uncertainty of parameters and interval variables,a probability interval mixed uncertain model is obtained.The constrained optimization problem is transformed into unconstrained optimization problem by using augmented Lagrange multiplier method.The paper tries to improve the convergence and stability of the algorithm when the limit state surface is highly nonlinear.The main contents of this paper are given follows:1.Aiming at the low computational efficiency of traditional reliability optimization algorithm,a structural reliability analysis method based on Augmented Lagrange multiplier method and PSO algorithm is proposed.In this method,a probability interval mixed uncertainty model is obtained by introducing parameter uncertainty and interval variables.The constrained optimization problem is transformed into an unconstrained optimization problem by using the augmented Lagrange multiplier method,and then the structural reliability is analyzed.Numerical examples and engineering examples show that the algorithm has good convergence and high computational efficiency for linear and nonlinear function functions.2.In order to solve the problem of high nonlinear limit state equation using traditional reliability optimization algorithm,a structural reliability analysis method based on Augmented Lagrange multiplier method and CS algorithm is proposed.The algorithm makes full use of the high efficiency of CS algorithm to solve the probabilistic interval mixed uncertain unconstrained optimization model,and obtains the optimal reliability index.Numerical examples and engineering examples show that the proposed algorithm has good stability and effectiveness in the calculation of structural reliability analysis problems.3.In view of the low efficiency of traditional reliability optimization algorithm in dealing with the reliability analysis of complex mechanical structures and the ignorance of correlation and uncertainty,a structural reliability analysis method based on Augmented Lagrange multiplier method and DE algorithm is proposed.This method takes the reliability index as the optimization objective function,establishes the mathematical model of reliability index,takes the limit equation as the constraint condition,and makes full use of the augmented Lagrange multiplier method to transform the constrained optimization problem into the unconstrained optimization problem,and uses the de algorithm to solve the structural reliability analysis problem.Numerical and engineering examples verify the stability and robustness of the proposed algorithm in solving complex reliability problems.