Optimization Method Based On Kriging Model For High-speed Rail Bogie Design

In some complex but important mechanical structures(such as high-speed rail bogies),various uncertain variables are involved,and the relationship between these variables and the response of actual engineering may be complicated.When optimizing design and reliability analysis,it is often necessary to use real physical model experiments or finite element analysis models to get the real response value,which is too inefficient and requires high cost.If the proxy model method is used,the required response value can be obtained by approximately fitting the relationship model between the actual input and output.Compared with the real physical model experiment and finite element model analysis,the analysis and calculation time can be greatly reduced and the time cost can be saved.Therefore,based on the Kriging proxy model,this thesis studies the optimization design and reliability analysis methods combined with the adaptive Kriging model,aiming at improving the efficiency and accuracy of the corresponding calculation methods,and applies these methods to the optimization design and fatigue reliability analysis of high-speed rail bogie frame.The main research contents of this thesis are as follows:(1)Introduce the basic principle of the Kriging model and the construction process of the adaptive Kriging model.Using Python code to realize the process of adaptive Kriging algorithm and through several numerical examples,the number of sample points required by the Kriging model combined with different sequential sampling methods,as well as the accuracy and efficiency of the obtained fitting model are compared.The comparison results show that Kriging model combined with sequential sampling methods has good calculation accuracy,but the number of samples required is still redundant.(2)Aiming at the shortcomings of the above sequential sampling Kriging method,an adaptive Kriging model method combined with global region division is proposed and applied to the multi-objective optimization design of high-speed rail bogies.Finally,the weight of the high-speed rail bogie frame is reduced by 16.483kg,with a reduction rate of 15.13%.Then the finite element model of high-speed rail bogie is used to evaluate the optimization results of this method.The evaluation results show that this method is effective and the error is less than 2%,which shows that this method can also be applied to complex engineering problems.(3)The adaptive Kriging model method combined with global division is applied to the reliability analysis of high-speed rail bogie frame.Compared with the traditional reliability analysis method,the error between the failure probability obtained by the constructed proxy model and the real value simulated by Monte Carlo method is less than 1%,but compared with the traditional reliability analysis method,the efficiency of analysis and calculation is improved by 33.51%.Based on the Kriging proxy model,this thesis studies the abovementioned optimization design method and structural reliability analysis method combined with Kriging model,which reduces the number of operations of the structural model,ensures the accuracy of the fitted model and improves the calculation efficiency of the corresponding algorithm.