基于代理模型的低速电动车车架多目标优化

Multi-objective optimization of the low-speed electric vehicle frame based on the surrogate model

为进行低速电动车车架的轻量化研究,采用有限元分析软件HyperMesh建立低速电动车车架有限元模型,并进行模态、强度和弯扭刚度分析。结合径向基神经网络(RBF)代理模型和粒子群优化算法,综合考虑质量、模态、强度和弯扭刚度等性能指标,对低速电动车车架进行轻量化多目标优化。结果表明,低速电动车车架质量降低了27.6 kg,同时其1阶模态频率、强度和弯扭刚度均满足设计要求,取得了较好的轻量化效果。

In order to explore the lightweight design of the low-speed electric vehicle frame,a finite-element model is worked out by HyperMesh;such factors as the mode,strength,as well as bending and twist stiffness of the frame are calculated and analyzed.Then,the RBF surrogate model and the algorithm of particle swarm optimization are employed for the multi-objective optimization of the frame,with such factors as mass,modality,strength as well as bending and twist stiffness taken into account.The results indicate that the optimized frame achieves a weight reduction of 27.6 kg,and the index meets the design requirements of first-order modal frequency,strength,as well as bending and twist stiffness.The proposed method proves effective for the lightweight design of the low-speed electric vehicle frame.