Abstract: In the modal analysis of vehicle frames using the finite element method, discrepancies between finite element calculation results and experimental test results inevitably arise due to factors such as uncertainty in material properties, simplifications in modeling and boundary conditions. This paper takes a newly developed special-purpose vehicle frame as the research object, establishes a fully parametric model, and employs a combined zero-order and first-order optimization method to update the finite element model of the frame based on experimental data. Taking the sum of frequency errors at each order as the objective function, material density, Young's modulus, and wall thickness are selected as design variables, while model mass error, individual frequency errors at each order, and modal assurance criterion are used as state variables for iterative computation. The results show that the objective function value decreases from 39.68% to 8.33%, the maximum single-order frequency error decreases from 6.03% to 2.13%, and the minimum modal assurance criterion (MAC) value corresponding to each order mode is 0.955. The updated frame model can be used for subsequent dynamic studies.

Key words: vehicle frame; zero-order optimization; first-order optimization; modal analysis; model updating

摘要： 在使用有限元法对车架进行模态分析时，由于材料特性的不确定性、模型和边界条件 简化等因素，使得有限元计算结果与试验测试结果存在一定的差距。文章以某新研型号特种 车车架为研究对象，建立全参数化模型，联合零阶与一阶优化方法，基于试验数据对车架有 限元模型进行修正，以各阶频率误差总和为目标函数，选择材料密度、杨氏模量及壁厚作为 设计变量，模型质量误差、各阶频率误差和模态置信度作为状态变量进行迭代运算，最终目 标函数的大小由 39.68%下降至 8.33%，单阶最大频率误差由 6.03%下降至 2.13%，各阶模态 对应的模态置信准则（MAC）值最小为 0.955，得到的车架模型可用于后续的动力学研究。

关键词: 车架；零阶优化；一阶优化；模态分析；模型修正