Reliabillity-based Structural Optimazation Based On Kriging Method

In the structural reliability analysis,the Kriging method is one of the common methods in engineering practice.The paper carries out the research on the Kriging method and improves the classical Kriging method,and applies the improved Kriging method to the reliability-based structural optimization.The main work of this paper is as following:1)Improve Kriging methodThe selection of samples has a great influence on the accuracy of Kriging model,a large number of samples will lead to the low efficiency,and few samples shouldn't ensure the accuracy of prediction.In this paper,an improved Kriging method is proposed based on gradient projection sampling.The improved method can make full use of the advantages of gradient projection technique,and can reasonably select the sample points for constructing Kriging model,which can solve the problem that the limit state function has one or more MPPs(Most Probable Points).2)Reliability-based structural optimizationIn this paper,the improved Kriging method is applied to reliability-based structural optimization design.A large number of examples show that the method has good efficiency and accuracy,and can greatly improve the efficiency of reliability-based structural optimization.3)ConclusionsThe numerical example results of structural reliability analysis show the improved Kriging method can has satisfactory accuracy for the MPP,reliability index and failure probability.Compared with the classical reliability analysis methods,the number of sample points is obviously reduced and the computational efficiency is improved.The reliability-based structural optimization example shows that the improved Kriging method can improve the optimization efficiency,obtain satisfactory results and have good robustness.