基于多目标可靠性优化平台的车身前部结构设计

Front Body Structure Design Based on Multi-Objective Reliability Optimization Platform

为了在保证精度的同时考虑不确定性因素高效开展车身结构优化设计,建立了多目标可靠性优化平台。采用径向基函数(RBF)代理模型代替仿真模型,基于概率充分因子法将概率约束转化为单一确定性约束,选取自适应加权多目标优化算法高效获取帕累托最优前沿(POF),基于智能补点方法生成补充样本点持续更新RBF模型,提升优化精度直至收敛。通过工字梁设计实例验证了该优化平台的有效性,并基于优化平台进行了某前部车身结构的多目标可靠性优化,结果表明,优化效果显著。

To efficiently perform optimization design for vehicle body structure in consideration of uncertainty factors while ensuring accuracy,a Multi-Objective Reliability Optimization(MORO)platform is established.The initial Radial Basis Function(RBF)model is built to substitute the simulation model and probabilistic constraints are converted into a single deterministic one with probabilistic sufficiency factor method.An adaptive weighted multiple-objective optimization algorithm is adopted to efficiently capture the Pareto Optimal Front(POF)and new samples are generated by an intelligent sampling approach to continuously update RBFs for optimization accuracy consideration.This updating iteration runs until MORO is converged.The proposed platform is proved to be effective by an I-beam design example and MORO for a front vehicle body is thus carried out on this platform,which demonstrates an obvious optimization.