| EMU bogie is a key component to ensure the smooth,stable,reliable and safe operation.In the traditional design,the contradictory problems of the bogie components is how to achieve the robustness and reliability of multi-quality characteristics,reduce the cost of production and improve the quality of production.Aiming at the problem,the multi-response robust optimization method,which can reduce the effects of controllable and uncontrollable factors on quality and effectively improve the quality,is introduced in the design of EMU key component.This method has got much attention and research in the academic circles at home and abroad.Based on these,the application of the multi-response robust optimization design method in the bogie components is studied and discussed.The main work is as follows.(1)Multi-response robust optimization design of bogie traction device.During the optimum design,the quality characteristics of bogie traction device involving quality(volume),longitudinal stiffness,and safety coefficient are affected by the value of controllable factors,such as structure size and tensile strength,and other uncontrollable factors,such as density,elastic modulus and so on.It is difficult to eliminate and reduce the influence of these factors,and to achieve the robustness of multiple quality characteristics as far as possible.The robustness of the traction device quality characteristics is realized by orthogonal experimental design and signal-to-noise tool and the designer’s preference information is considered by using physical planning method,then the preference function is constructed.The multi-response robust optimization design model of traction device is built by properly translating into a mathematical model to represent the engineering problems.Based on the combination of the signal-to-noise ratio and the physical programming method,the multi-response robust optimization design method is easy applied in engineering practice widely.It is easy and wide for the multi-response robust optimization design method that applied in engineering practice.(2)Multi-response robust optimization design of traction device based on stochastic model.In the multi-response robust optimization design,the random design variables of bogie traction device are considered by using the normal distribution,and the robustness of objective function and constraint conditions is achieved by minimizing the expected loss function and the constraint feasibility criterion,respectively,then the multi-response robust optimization model is established based on the stochastic model.The model is calculated with the discrete algorithm,which is combined with Monte Carlo sampling method and NSGA-Ⅱ,and the resulting stochastic robust optimization solution are compared and analyzed with the multi-response robust optimization solutions,then the robust optimal design solution,which has a smaller variance then instance,is obtained.This method can reasonably describe the engineering application of random uncertainty.(3)Multi-disciplinary robust optimization design of bogie traction gearbox.The three disciplines of the traction gearbox involving the structural design,mechanical design and the requirements of vibration noise are considered,it is necessary to combine the multi-disciplinary optimization design and the multi-response robust optimization design.A target mean and variance of the dual response surface model is created by using Box-Behnken design,and the fitting function mean and variance is got to ensure the robustness of the various disciplines through the dual response surface method.Then the layer system of multi-disciplinary robust optimization model is built by the application of collaborative optimization and the integrate calculation of each subject.Finally,with the bogie traction gearbox as an example,the multidisciplinary robust optimization design model is established,and the optimization model is calculated on the ISIGHT software,which is multi-disciplinary design optimization software.Then the multi-disciplinary robust optimal solutions and the conventional robust optimization design solutions are compared and analyzed.The results indicate that this method is feasibility and has more practical value in engineering application. |
