基于T-S模糊模型的车辆电液悬架系统H_∞控制  

Takagi-Sugeno Fuzzy-model-based Automobile Electrohydraulic Active Suspension H-infinity Control

作　　者：余曼[1,2] 周辰雨 魏朗[1] 赵轩[1] 余强[1] YU Man;ZHOU Chen-yu;WEI Lang;ZHAO Xuan;YU Qiang(School of Automobile,Chang＇an University,Xi＇an 710064,Shaanxi,China;Xi＇an Automobile Maintenance Industry Management Office,Xi＇an 710061,Shaanxi,China)

机构地区：[1]长安大学汽车学院,陕西西安710064 [2]西安市汽车维修行业管理处,陕西西安710061

出　　处：《中国公路学报》2018年第8期205-217,共13页China Journal of Highway and Transport

基　　金：国家重点研发计划项目(2017YFC0803904);国家自然科学基金项目(51507013);中国博士后基金项目(2018T111006;2017M613034);陕西省博士后基金项目(2017BSHEDZZ36);陕西省重点产业创新链(群)项目(2018ZDCXL-GY-05-03-01);陕西省重点研发计划项目(重点项目)(2018ZDXM-GY-082);陕西省创新人才推进计划-青年科技新星项目(2018KJXX-005)

摘　　要：为了进一步降低主动悬架作动器输出力并优化控制系统鲁棒性,建立车辆7自由度整车模型,采用Takagi-Sugeno(T-S)模糊建模技术,设计主动悬架外环H_∞控制器,从而根据路面输入调节主动悬架性能,提升作动器能效。通过构建一个包括控制器稳定性分析、悬架运动空间及力的限值问题的线性矩阵不等式组,将控制器的优化问题转换为此线性不等式组的求解问题,并结合并行分配补偿控制技术,得到此控制器状态反馈系数。针对系统不确定性参数,内环采用自适应鲁棒控制方法,提升控制力的跟踪性能。通过对不同路面轮廓激励工况、交叉轴双轮激励工况以及控制力跟踪性能进行仿真试验,分析被动悬架和主动悬架性能评价指标,并对其作动器输出力进行对比研究。研究结果表明:在小激励下,基于T-S模糊模型的H_∞控制主动悬架相比被动悬架,各车轮处加速度均方根值可降低80%以上,与最优控制相比可降低47%以上;而在大激励时,虽然其加速度均方根值有所上升,但其悬架动挠度峰值较被动悬架有所下降;通过路面交叉轴激励对比可以看出,针对整车平顺性参数,该方法可在路面小激励时较被动悬架降低质心、俯仰以及侧倾加速度均方根值达55%、83%以及90%以上;与反演作动器输出力及最优控制作动器输出力对比结果表明,该控制方法可有效降低主动悬架控制力峰值20%以上,并提升控制力的跟踪性能;基于T-S模糊模型的H_∞控制可以在保证车辆悬架性能的基础上有效降低系统能耗。To reduce the output force of active actuator and improve the robustness of a control system,an H_∞-outer-loop controller was built based on a 7-degree-of-freedom full car model of the Takagi-Sugeno（T-S）fuzzy modeling technique to adapt its strategy to different road disturbances and improve system efficiency.The resulting optimizing problem was transformed into a solution issue of linear matrix inequalities associated with system stability analysis,suspension stroke limit,and force constraints.By integrating these via parallel distributed compensation method,the controller gain was obtained.In response to the uncertain parameters,the adaptive robust control was employed in the inner loop to improve the force-tracking performance.Based on the simulation test under the condition of different road profiles,crossaxle double-road input,and control force tracking,the suspension performance indexes were analyzed and the control force was compared to backstepping and optimal control.The results show that the T-S fuzzy model based on the H_∞ control can reduce the root mean square（RMS）value of acceleration by more than 80% and 47% compared to passive suspension and optimal control,respectively.The RMS value of acceleration increases when under intensive disturbance with suspension dynamic deflection suppression.In cross-axle double-road input tests,the mean values of mass,pitch,and roll accelerations were reduced by more than 55%,83%,and 90%,respectively,compared to those of passive suspension under a small road disturbance.The peak control force was observed to reduce by more than 20%than that of the backstepping control and optimal control method,and the tracking accuracy was improved.These results show that the combination of the H_∞ control and adaptive robust control based on the T-S fuzzy model can reduce the system energy consumption with a guaranteed suspension performance.

关 键 词：汽车工程 主动悬架 T-S模糊建模 自适应鲁棒控制 不确定性参数 

分 类 号：U463.1[机械工程—车辆工程]