Numerical model and multi-objective optimization analysis of vehicle vibration

Abstract

It is crucial to conduct a study of vehicle ride comfort using a suitable physical model, and a precise and effective problem-solving method is necessary to describe possible engineering problems to obtain the best analysis of vehicle vibration based on the numerical model. This study establishes different types of vehicle models with different degrees of freedom (DOFs) that use different types of numerical methods. It is shown that results calculated using the Hamming and Runge-Kutta methods are nearly the same when the system has a small number of DOFs. However, when the number is larger, the Hamming method is more stable than other methods. The Hamming method is multi-step, with four orders of precision. The research results show that this method can solve the vehicle vibration problem. Orthogonal experiments and multi-objective optimization are introduced to analyze and optimize the vibration of the vehicle, and the effects of the parameters on the dynamic characteristics are investigated. The solution F ₁ (vertical acceleration root mean square of the vehicle) reduces by 0.0352 m/s², which is an improvement of 7.22%, and the solution F ₂ (dynamic load coefficient of the tire) reduces by 0.0225, which is an improvement of 6.82% after optimization. The study provides guidance for the analysis of vehicle ride comfort.

中文概要

目 的

通过采用不同数值方法求解不同的车辆动力学模 型, 为车辆动力学模型研究提供参考; 结合正交 试验和多目标优化算法来分析各个参数对车辆 性能的影响权重, 采用多目标优化算法进行车辆 动力学多目标优化分析, 为车辆的设计提供参考 依据。

创新点

研究不同数值方法的求解精度, 为车辆动力学求 解方法提供新途径; 采用正交试验设计研究车辆 各参数的影响权重, 为车辆设计提供参考; 采用 多目标优化算法设计车辆, 能兼顾车辆多个方面 的性能。

方 法

采用不同动力学求解算法、正交试验设计和多目标优化分析方法。

结 论

1. 基于不同数值求解算法的研究表明, Hamming 法要优于Newmark 法和有限差分法, 四阶 Hamming 法的精度不如龙格库塔法; 2. 正交试验 可得到各参数对车辆动力学的影响权重, 但忽略 了参数间的交互效应; 3. 经过多目标优化设计, 衡量车辆振动性能的两个指标分别减少了7.22% 和6.82%。