Robust Optimization Design Of Gearbox Housing For EMU

The gearbox housing of EMU is not only one of the important components of the train traction transmission system,but also the key part to ensure the safe operation of the train.With the continuous improvement of the running speed of EMU,the working environment of the gearbox housing is getting worse and worse.However,due to the restriction of manufacturing conditions,the gearbox housing is affected by many uncertain factors in the design and manufacturing,resulting in the fluctuation of its quality characteristics.Therefore,it is necessary to study the robustness of the quality characteristics of the gearbox housing to improve the robustness of its structural performance.This thesis will introduce the idea of robust design,and combine the computer simulation technology and intelligent optimization technology to carry out the robust optimization design of the gearbox housing,in order to improve the design quality of EMU gearbox housing.The main research work of this thesis is as follows:Firstly,the static strength and mode of EMU gearbox housing under extreme working conditions are simulated and analyzed to verify whether its structure meets the design requirements.In view of the random uncertainty of the design variables of the gearbox housing,the structural parameters that have a great impact on its structural strength are selected as the design variables by sensitivity analysis.The sensitivity index and the expected loss function are both minimized as the optimization objective,and the robust optimization model of gearbox housing is constructed.The robust optimal solution of the gearbox housing is obtained by using multi-population genetic algorithm.The design of the optimal solution not only improves the robustness of its structural performance,but also realizes the lightweight design.Secondly,in view of the multi-discipline involved in the design of gearbox housing and the interaction between disciplines,considering the random uncertainty of the design variables,the multi-disciplinary robust design model of the gearbox housing is constructed by combining the multi-disciplinary collaborative optimization method and the 6σ robust design theory.Among them,the structural weight,strength and modal disciplines of the gearbox housing are comprehensively considered.The weight is the system level optimization discipline,and the static strength and modal characteristics are the subsystem level optimization disciplines.The optimization consistency of various disciplines is coordinated through the system level consistency constraints.The two-layer multidisciplinary collaborative robust optimization model for gearbox housing is established,and the best robust design scheme is obtained by solving with the help of the Isight multidisciplinary optimization integration platform.Through the comparative analysis with the original design scheme,the feasibility of this method is verified,and it is proved to have certain engineering practical significance.Finally,in order to improve the structural reliability of the gearbox housing and ensure its robustness,the reliability robust design of the EMU gearbox housing is proposed.Taking the minimum reliability sensitivity as the optimization objective and the structural reliability as the constraint condition.The reliability robust optimization design model of gearbox housing is constructed by using artificial neural network.The reliability robust optimal solution of gearbox housing is obtained by using the optimization process of neural network.In addition,in view of the problem that the fitting accuracy of the traditional BP neural network is greatly affected by the initial weight values and threshold values,a BSO-BP neural network model is proposed to improve the accuracy of the neural network model.The results show that the optimization results effectively improve the reliability and robustness of the gearbox housing design scheme,and further expand the application of reliability robust design method in engineering practice.