东北大学学报(自然科学版) ›› 2025, Vol. 46 ›› Issue (4): 43-51.DOI: 10.12068/j.issn.1005-3026.2025.20230293

• 机械工程 • 上一篇    下一篇

考虑辅助部件的车辆动力总成系统建模方法与优化设计

吴杨俊1, 李振平2, 姚红良1, 韩圣东1   

  1. 1.东北大学 机械工程与自动化学院,辽宁 沈阳 110819
    2.中国北方车辆研究所,北京 100072
  • 收稿日期:2023-10-20 出版日期:2025-04-15 发布日期:2025-07-01
  • 作者简介:吴杨俊(1994—),男,江西抚州人,东北大学博士研究生
    姚红良(1979—),男,河北唐县人,东北大学教授,博士生导师.
  • 基金资助:
    中央高校基本科研业务费专项资金资助项目(N2303005)

Modeling and Optimization Design of Vehicle Powertrain System Considering Effect of Auxiliary Components

Yang-jun WU1, Zhen-ping LI2, Hong-liang YAO1, Sheng-dong HAN1   

  1. 1.School of Mechanical Engineering & Automation,Northeastern University,Shenyang 110819,China
    2.China North Vehicle Research Institute,Beijing 100072,China. Corresponding author: YAO Hong-liang,E-mail: hlyao@mail. neu. edu. cn
  • Received:2023-10-20 Online:2025-04-15 Published:2025-07-01

摘要:

基于神经网络参数识别法与子结构模态综合法,提出一种考虑辅助部件的动力总成非线性系统建模方法,并运用遗传算法对辅助部件的连接刚度及阻尼进行了多目标优化设计.首先,基于神经网络对动力总成系统模型进行拟合,并以试验模态参数为目标,运用遗传算法对动力总成系统的连接刚度及阻尼进行参数识别,结果显示仿真与试验的模态频率最大误差为-5.98%,模态阻尼比的最大误差为-15.72%.然后,运用子结构模态综合法对动力总成系统进行缩聚处理,并研究了辅助设备与发动机间的耦合振动影响情况.最后,以辅助部件的振动性能最优为目标对连接件刚度及阻尼参数进行多目标优化设计,优化后模型中冷器与空气滤清器的位移最大峰值分别较优化前下降了34.6%与4.61%.

关键词: 动力总成建模, 辅助部件, 神经网络, 参数识别, 子结构, 多目标优化

Abstract:

Based on the neural network parameter indentification method and component mode synthesis (CMS), a modeling approach for the nonlinear powertrain system is proposed to investigate the coupled vibrations of the engine and auxiliary components, and the multi-objective optimization design using genetic algorithms is applied to optimize the connection stiffness and damping. Firstly, a neural network-based model was employed to fit the dynamic model of the powertrain system. According to the experimental modal parameters, the genetic algorithms were applied to identify the connection stiffness and damping of the powertrain system. The results showed that the maximum discrepancies between simulated and experimental modal frequencies and damping ratios were -5.98% and -15.72%, respectively. Subsequently, the CMS is employed to reduce the degrees of freedom of the powertrain system, and the engine-equipment coupling vibration response is evaluated. Finally, a multi-objective optimization design was performed to achieve the optimal vibration performance of the auxiliary components. The maximum peak values displacement of the intercooler and air filter for the optimized model decreased by 34.6% and 4.61%, respectively, compared to the original ones.

Key words: powertrain modeling, auxiliary components, neural network, parameter identification, substructure, multi-objective optimization

中图分类号: