The Reliability-based Design Optimization Based On Probability And Interval Hybrid Model

Uncertainties exist widely in practical engineering problems.The method to deal with uncertainty is related to the safety of the structure.Reliability-based design optimization(RIBDO)can fully consider the influence of parameter uncertainty on the structure,so as to ensure the optimal solution to meet the reliability requirements,which plays an important role in the safety design of engineering structures and products.Most of the existing researches use probability model to describe the uncertain variables.However,to describe the uncertainty through probability model,it is necessary to master the accurate probability density distribution function of variables.In practical engineering problems,it is often difficult to be satisfied.So it is not suitable for many engineering design problems to build reliability-based design optimization problems only through probability model.In order to make up for the limitation of probability model,the non-probabilistic convex set model can be used to supplement the probability model.Interval model is one of the non-probabilistic convex set model,which is only need to determine the upper and lower limits instead of accurate density distribution function.Therefore,this paper explores and studies the reliability-based design optimization based on probability interval hybrid model,and the main research contents and results are as follows:1.A method of reliability-based design optimization based on probability interval hybrid model is proposed.By using the nested decoupling strategy and the maximum entropy principle method,the three-layer reliability-based design optimization problem is equivalent to the single-layer,which improves the calculation efficiency while ensuring the accuracy of the maximum failure probability with hybrid model constraint function.2.A hybrid reliability-based design optimization method based on local encryption approximation model technology is proposed.The radial basis function is used to establish the approximate model of the objective function and the constraint function of the optimization.Therefor the complex and time-consuming numerical model is replaced.According to the results,the error of the approximate model and the numerical model is compared,and the local encryption method is used to achieve higher accuracy,which further improves the efficiency and accuracy of the optimization.3.A software platform for reliability-based design optimization based on probability interval hybrid model is developed based on Visual Basic 6.0.The software platform is applied to the optimization of vehicle crashworthiness,and the automation of optimization design is realized.