【摘要】 为了提升整车路噪、平顺性以及疲劳耐久综合性能，构造了序列近似多目标优化方法，结合虚拟试验场(virtual proving ground, VPG)技术对轮胎参数进行了优化设计。采用径向基函数代理模型建立了CDTire轮胎模型关键物理参数与整车多学科性能VPG仿真输出响应的显式数学关系，基于多目标粒子群优化算法与自适应补点方法持续更新代理模型迭代求解逐步逼近真实Pareto前沿。通过设计变量分类建立2步法优化模型，求解匹配轮胎模型参数后，车辆关键测点的全频段噪声均方根、最大振动加速度均值及疲劳损伤均值分别降低了1.6%,10.4%及3.8%,优化效果显著。更多还原

【Abstract】 Here, to improve comprehensive performance of road noise, ride comfort, and fatigue durability of entire vehicle, a sequence approximation multi-objective optimization method was constructed, combined with virtual testing field technology, tire parameters were designed and optimized. Radial basis function surrogate model was used to establish explicit mathematical relations between key physical parameters of CDTire tire model and entire vehicle multidisciplinary performance VPG simulation output responses. Based on the multi-objective particle swarm optimization algorithm and the adaptive patching method, the surrogate model was continuously updated and solved iteratively to gradually approach the real Pareto frontier. A two-step optimization model was established through classification of design parameters. After solving matched tire model parameters, it was shown that RMS of noise over full frequency range, mean maximum vibration acceleration, and mean fatigue damage of vehicle’s key measured points decrease by 1.6%, 10.4%, and 3.8%, respectively, so the optimization effect is significant.更多还原