汽车半主动磁流变悬架的自适应相对控制研究  

Adaptive Relative Control Method for Automotive Semi-active Suspension with Magnetorheological Damper

作　　者：杨建伟[1] 孙守光[2] 刘海波[3] 

机构地区：[1]北京建筑工程学院机电与汽车工程学院,北京100044 [2]北京交通大学机械与电子控制工程学院.北京100044 [3]太原科技大学机械电子工程学院,太原030024

出　　处：《系统仿真学报》2011年第9期1951-1955,共5页Journal of System Simulation

基　　金：北京市属高等学校人才强教计划项目(PHR201106125);北京市优秀人才培养资助项目(2010D005017000007);山西省青年学术带头人项目(20081066);北京建筑工程学院博士启动基金(101001704)

摘　　要：为了提高汽车的运行平顺性和操纵稳定性,以磁流变减振器为控制对象,提出了采用自适应神经网络相对控制方法的半主动悬架系统。在试验测试和理论分析的基础上,通过数据拟合得到磁流变减振器阻尼力的非线性Bingham模型,建立了基于该磁流变减振器的半主动悬架模型,并用该模型进行了自适应神经网络相对控制方法的研究。以模拟道路谱作为输入,以车身俯仰角加速度和车辆悬架前、后侧弹簧的垂向变形量之差作为控制量,把车身质心垂向加速度、前后悬架动行程作为评价指标来进行仿真研究。仿真结果表明,设计的半主动悬架与被动悬架相比,其平顺性与稳定性均得到了良好的改善,且对运行工况的改变有一定的适应性,对系统参数摄动也具有很强的鲁棒性。In order to improve automotive＇s riding performance and handling stability,an adaptive Neural Network（NN） relative control algorithm was applied to the semi-active suspension system and the controlled member is the Magnetorheological（MR） damper.Based on experimental test and academic analysis,a nonlinear Bingham model for MR damper damping force was obtained through data fitting,and an automotive semi-active suspension model with MR damper was built,which could be used to study the adaptative NN relative control method.The input of the model is a simulative road spectrum,and the controlled variables are the pitch angle acceleration and the vertical deformation＇s difference of the front and rear spring,and the evaluating indicators are the vertical centroidal acceleration of car body and the working space of the front and rear suspension.Then the simulation was carried on.Comparing with the passive suspension,the riding performance and the handling stability of designed semi-active suspension were improved,and the semi-active suspension system has self-adaptability to the working condition and better robust to the perturbation of the system parameters.

关 键 词：车辆半主动悬架 磁流变减振器 自适应神经网络相对控制 试验研究 数值仿真 

分 类 号：U463-33[机械工程—车辆工程]