基于磁流变阻尼器的车辆半主动悬架GA-LQR的研究  

Research on GA-LQR of automotive semi-active suspension based on MRD

作　　者：邵帅 胡国良 顾瑞恒 杨程 涂渝 SHAO Shuai;HU Guoliang;GU Ruiheng;YANG Cheng;TU Yu(Key Laboratory of Conveyance and Equipment,Ministry of Education,East China Jiaotong University,Nanchang 330013,China)

机构地区：[1]华东交通大学载运工具与装备教育部重点实验室,南昌330013

出　　处：《现代制造工程》2021年第11期1-9,共9页Modern Manufacturing Engineering

基　　金：国家自然科学基金项目(51765016)。

摘　　要：传统的线性二次型最优控制(LQR)策略可提升车辆乘坐舒适性及操作稳定性,因此广泛应用在车辆半主动悬架中。对车辆半主动悬架用磁流变阻尼器(MRD)进行阻尼特性试验,并采用Bouc-Wen模型建立MRD的力学模型。为确定Bouc-Wen模型中的未知参数,利用遗传算法(GA)对其进行参数辨识。结合所辨识的Bouc-Wen模型,建立1/4车辆磁流变半主动悬架模型。针对LQR控制器存在的加权矩阵Q和R的取值易受主观因素影响的问题,设计了一种基于遗传算法的线性二次型最优控制(GA-LQR)控制器。以随机路面激励作为输入,分别对被动悬架、基于LQR的磁流变半主动悬架和基于GA-LQR的磁流变半主动悬架进行仿真分析,结果表明,基于GA-LQR的磁流变半主动悬架具有更好的乘坐舒适性及操作稳定性。Linear Quadratic Regulator(LQR)strategy is widely applied to automotive semi-active suspension due to its performance in improving the riding comfort and operation stability of the vehicle.The damping characteristics of the designed Magnetorheological Damper(MRD)were tested,and the Bouc-Wen model was applied to establish the mechanical model of MRD.To ensure unknown parameters in the Bouc-Wen model,Genetic Algorithm(GA)was used to identify those parameters.Combined with the identified Bouc-Wen model,the quarter of the automotive magnetorheological semi-active suspension system model was constructed.The values of the weighted matrixes Q and R of the LQR controller were significantly influenced by subjective factors,hence,an LQR based on GA controller(GA-LQR controller)was designed.Under the excitation of a random road,the comparison between the conventional passive suspension and the magnetorheological semi-active suspension applied with LQR control and applied with GA-LQR control,were theoretically conducted and analyzed.The results indicate that the proposed control strategy can shares a promising riding comfort and operation stability.

关 键 词：磁流变阻尼器 半主动悬架系统 BOUC-WEN模型 参数辨识 遗传算法 GA-LQR控制器 

分 类 号：U463.33[机械工程—车辆工程]