部分可观测车辆系统非线性随机振动的最优控制  

Optimal Control of Nonlinear Stochastic Vibration of Partially Observable Vehicle Systems

作　　者：张巍[1] 应祖光[2] 颜光锋 ZHANG Wei;YING Zuguang;YAN Guangfeng(Laboratory Center,School of Economics and Management,Zhejiang Sci-Tech University,Hangzhou 310018,China;Department of Mechanics,School of Aeronautics and Astronautics,Zhejiang University,Hangzhou 310027,China)

机构地区：[1]浙江理工大学经济管理学院实验中心,杭州310018 [2]浙江大学航空航天学院力学系,杭州310027

出　　处：《噪声与振动控制》2018年第6期7-11,共5页Noise and Vibration Control

基　　金：国家自然科学基金资助项目(11572279)

摘　　要：车辆运行过程的随机振动水平是评估其动力学性能的重要指标,该振动对于车载器件能否正常工作具有极其重要的影响,因此必需进行车辆随机振动控制。重型多轴车辆受空间限制其悬架采用可转动的斜杆支承,且控制器如磁流变阻尼器也斜向安装在悬架与车轮之间,导致系统呈现几何非线性,其非线性随机振动控制方法与效果完全不同于普通车辆。同时由于不可避免的观测噪声,导致出现部分可观斜杆支承车辆系统的非线性随机控制新问题。考虑车体与车轮的垂直耦合运动及斜支承杆的转动,用拉格朗日方程建立车辆控制系统模型的运动微分方程,转化为非线性的耦合振动方程,同时建立包含测量噪声的系统观测方程,构成一个部分可观系统的非线性随机最优控制问题;根据推广的Kalman滤波方法得到关于估计状态的非线性随机系统方程,再根据随机动态规划原理建立动态规划方程,结合控制力的有界性,得到基于系统估计状态的最优有界控制律;通过受控与未控系统响应统计的比较评估控制效果,数值计算结果表明该控制策略可有效降低具有观测噪声的采用斜杆支承与控制车辆系统在随机路面激励下的非线性随机振动,并对于不同观测系数具有一定的鲁棒性。Stochastic vibration level of vehicles during driving is an important criterion to evaluate their dynamic performance. The stochastic vibration has a great impact on the working condition of vehicle devices, and therefore must be controlled. The suspension of heavy-duty multi-axle vehicles is supported aslant by rotatable levers because of the space restriction. The control devices such as MR dampers axe also installed aslant between the suspension and wheels, which leads to the geometric nonlinearity of vehicle systems. The control strategy and effectiveness of the nonlinear stochastic vibration vehicles axe totally different from those of ordinary vehicles. Furthermore, because of the inevitable state measurement noise, the rotatable-inclined-support vehicle control becomes a new problem of partially observable nonlinear stochastic control. In this paper, the vertical coupling motion between the suspension and wheels and the rotation of the inclined support levers are considered. The differential equations of motion of the vehicle system are derived according to the Lagrangian equations, and then transformed into nonlinear coupled vibration equations. Meanwhile, the state observation equations with measurement noise are given. These equations can describe the partially observable nonlinear stochastic control problem. The nonlinear stochastic system equation for estimation states is obtained based on the extended Kalman filter. Then, the dynamic programming equation is obtained based on the stochastic dynamical programming principle. The optimal bounded control law for the estimation states is determined by the programming equation with the boundedness of control forces. The effectiveness of the control strategy is evaluated by comparing the response statistics of controlled and uncontrolled systems. Numerical results show that the proposed control strategy can effectively mitigate the nonlinear stochastic vibration of the rotatable-inclined-support vehicle system with noise observation under random road excitat

关 键 词：振动与波 随机振动 非线性车辆系统 部分可观测 最优控制 

分 类 号：O324[理学—一般力学与力学基础] O328[理学—力学]