减振器动态特性集总参数-神经网络联合模型  

Lumped Parameter-Neural Network Hybrid Model for Dynamic Characterization of Automotive Dampers

作　　者：刘雪莱 刘子谦 郑益谦 王强 李涛[3] 上官文斌[3] LIU Xuelai;LIU Ziqian;ZHENG Yiqian;WANG Qiang;LI Tao;SHANGGUAN Wenbin(Chongqing Qingshan Industry Co.,Ltd.Chongqing,402761,China;College of Packaging Engineering,Jinan University Zhuhai,519070,China;School of Mechanical and Automotive Engineering,South China University of Technology Guangzhou,510640,China)

机构地区：[1]重庆青山工业有限责任公司,重庆402761 [2]暨南大学包装工程学院,珠海519070 [3]华南理工大学机械与汽车工程学院,广州510640

出　　处：《振动.测试与诊断》2025年第2期354-358,415,416,共7页Journal of Vibration,Measurement & Diagnosis

基　　金：国家自然科学基金资助项目(52405106)。

摘　　要：为了描述减振器的动态特性,将集总参数模型和神经网络模型进行结合,提出了一种能够精确描述减振器非线性动态特性的集总参数-神经网络联合模型。首先,在减振器测试台架上对一个减振器的动态特性进行测量,激励工况包括谐波激励和瞬态激励两大类,测试得到减振器的动态力响应;其次,提出了联合模型的集总参数建模方法与参数辨识方法,并使用谐波激励下的测试数据完成了模型参数辨识;然后,利用神经网络模型补偿减振器的力-速度非线性特征;最后,使用联合模型计算在瞬态激励下减振器的输出力,并与集总参数模型和神经网络模型的计算结果以及测试结果进行了对比。结果表明,所提出的减振器联合模型具有较高的准确性,能够有效补偿集总参数模型的非线性误差。To characterize the dynamic behavior of automotive dampers,a hybrid model integrating a lumped parameter model with a neural network is proposed to accurately describe their nonlinear dynamic characteristics.First,dynamic tests are conducted on a damper test rig under harmonic and transient excitations to measure the force response.Next,a parameter identification framework for the hybrid model is developed,and its parameters identification are calibrated using harmonic excitation test data.A neural network model is then introduced to compensate for the nonlinear force-velocity relationships that the lumped parameter model fails to capture.Finally,the hybrid model is used to calculate the output force of damper under transient excitations,and the results are compared with those of the lumped parameter model.The results indicate that the proposed hybrid model of dampers achieves high accuracy and can effectively mitigates the nonlinear modeling errors of lumped parameter models.

关 键 词：减振器 联合模型 神经网络模型 瞬态激励 

分 类 号：TH113[机械工程—机械设计及理论] U461[机械工程—车辆工程]