Abstract: Generally, the inertia properties of a powertrain mounting system (PMS) are determined by measuring, and the measurement error is less than 3 %. The PMS frequency and decoupling ratio are closely related to the inertia properties, mount stiffness, and mount position, thus the actual PMS frequency and decoupling ratio will not be consistent with their designed values. In this paper, the inertia properties were described with uniformly distributed random variables, the mount stiffness parameters were selected as design variables, and a robust optimization method was presented in order to desensitize the PMS frequency and decoupling ratio to the inertia properties. The optimization results of a transversely mounted PMS show that the robustness of the frequency, decoupling ratio, and frequency separation can be effectively improved compared with that of the traditional deterministic optimization method.

Key words: vibration and wave, powertrain mounting system, natural frequency, decoupling ratio, robustness optimization, [σ] level

摘要： 汽车动力总成的惯性参数（转动惯量和惯性积）通常采用实验方法进行测量，测量误差一般不超过3 %。动力总成悬置系统的固有特性与动力总成惯性参数、悬置刚度、位置、角度密切相关，从而悬置系统的固有频率和解耦率的理论设计值与其真实值之间必然存在一定程度偏差。采用均匀分布随机变量描述惯性参数，以悬置刚度为优化设计变量，提出稳健优化模型对某轿车悬置系统固有特性进行稳健优化。优化结果表明，与确定性优化方法相比，稳健优化方法可以较大幅度地提高频率、解耦率、频率间隔的稳健性。

关键词: 振动与波, 动力总成悬置系统, 固有频率, 解耦率, 稳健优化, [σ]水平