节点文献

基于微分几何解耦的汽车主动悬架控制算法研究

Research on Control Algorithm of Vehicle Active Suspension Based on Differential Geometry Decoupling Theory

【作者】 陈建国

【导师】 程军圣;

【作者基本信息】 湖南大学 , 机械工程, 2013, 博士

【摘要】 车辆悬架系统不仅影响着车辆的平顺性,而且影响着车辆操纵稳定性。在平顺性方面,由于四个车轮受到路面的激励,车身的振动是各车轮引起振动的综合结果。在操稳性方面,当车辆转向时,由于前轮转向角的影响,车辆车身的侧倾会加剧;当车辆处于紧急转向工况,由于纵向及横向载荷转移,四个悬架的垂向力会发生改变,从而使各轮胎的抓地性发生改变,对紧急转向时车辆的稳定性造成影响。车辆在不同工况下,控制的目的是不同的。车辆悬架控制系统是一MIMO的非线性控制系统。车辆主动悬架的控制问题,本质上是MIMO系统的控制分配问题,即数学上著名的Morgan问题。微分几何解耦控制理论是解决非线性系统控制问题的有力工具,可以实现对主动悬架控制力的合理分配。本课题针对车辆的不同工况,利用微分几何解耦控制理论,对车辆主动悬架控制算法进行研究。(1)车辆在直线工况下,主动悬架控制的目的是提高车辆平顺性。提出了不包括阻尼器解耦的EDDC控制算法。悬架系统经过解耦,车身的垂向、侧倾及俯仰运动各自独立。悬架系统被解耦成了多个独立的线性子系统。通过线性子系统的极点配置,车身的垂向、俯仰及侧倾振动大幅衰减,时域仿真、频域仿真及基于dSPACE硬件实时仿真验证了解耦算法的有效性。(2)车辆在转向工况下,车身的侧倾控制是主动悬架控制的重点。通过符号运算,在车身侧倾表达式中显式表达了与车辆前轮转角相关的项,明确了前轮转角对侧倾造成的影响;利用EDDC解耦算法,对车身的姿态进行解耦,使车身的各向运动相互独立,并对前轮转角对车身侧倾的干扰项通过解耦来进行抑制;分别在前轮转角输入为阶跃信号及正弦信号的情况下进行了仿真,验证解耦算法的有效性。(3)对于紧急转向工况,车辆的横向稳定性是车辆主动悬架控制的重点。建立了包括Burckhardt非线性轮胎模型的整车横向稳定性控制的数学模型,并把该模型表达成仿射非线性的形式,为微分几何解耦控制奠定了基础。(4)提出了基于PI制动与主动悬架解耦的车辆横向稳定性联合控制算法。采用PI制动来改善紧急转向时车辆的横摆;以主动悬架的主动作动器为输入,利用解耦算法来调节主动悬架的垂向载荷,改善轮胎的抓地性,从而调节车辆的侧向力,达到减小车身侧偏角的目的。对车辆在干燥路面及湿滑路面下,分别进行了阶跃输入及单移线输入仿真,验证了PI制动与主动悬架解耦联合控制算法的有效性。

【Abstract】 The vehicle suspension system influences not only harshness, but also handling stability of the vehicle. In the aspect of harshness, four wheels undergo the excitation from road; the vibration of the vehicle body is the resultant which all the wheels cause. In the aspect of handling stability, due to the influence of the front wheel angle, the roll motion of the vehicle body gets aggravated in steering condition. When the vehicle is in an emergence steering condition, since the longitude and the lateral load transferring, the vertical force of four suspensions alter, so the holding ability of the tires changes correspondingly and that influences the handling stability of the vehicle. The control objectives of the vehicle in different conditions are different. The vehicle suspension control system is a MIMO nonlinear system. The control of vehicle active suspension is essentially the control allocation of a MIMO system, namely the well-known Morgan problem in mathematics. The differential geometry control theory is a powerful tool to solve the control problem of the nonlinear systems and can realize the reasonable allocation of the active suspension force.Considering the different working conditions of vehicles, the decoupling control algorithms on the active suspension have been researched by using differential geometry theory in this project.(1) In straight line driving condition, the control objective of the active suspension is to improve the harshness of the vehicle. The EDDC (excluding damper decoupling control) algorithm is proposed. The vertical motion, pitch motion and roll motion of the vehicle body are independent of each other by decoupling, and the suspension system is separated into multi independent linear subsystems. Through the pole assigned of the linear subsystem, the vertical motion, pitch motion and roll motion of the vehicle body attenuate greatly. The simulation of the time domain, the frequency domain, and the real time simulation based on dSPACE hardware verify the effectiveness of the decoupling control algorithm.(2) In steering condition, the control objective of the active suspension is the roll control of the vehicle body. Through the symbol computing, the items associated with the front wheel corner are expressed explicitly in the roll equation of the vehicle body and makes the influence on the roll motion which caused by the front wheel corner clear. By using the EDDC decoupling algorithm, the posture of the vehicle body is decoupled and makes the motion on each direction independent of each other, and the interference item to the roll motion caused by the front wheel corner is restrained by decoupling. The simulation results of the step and sine input signal verify the effectiveness of the decoupling algorithm.(3) In emergency steering condition, the lateral stability is the key objective of the active suspension control. The lateral stability control mathematic model of the whole vehicle which including the Burckhardt nonlinear tire model is established, and the model is expressed in affine nonlinear form. The above lays the foundation for differential geometry decoupling control.(4) The union algorithm of the vehicle lateral stability control based on the PI braking and the active suspension decoupling is proposed. Using PI braking strategy the yaw motion of the vehicle in emergency steering condition is improved. Set the actors of the active suspensions as input, using the decoupling algorithm to adjust the vertical load, improve the holding ability of the tires thus its lateral force is adjusted correspondingly, as a result, the aim to decrease the side slip angle of the vehicle is achieved. Simulations are carried out both on the dry and the wet surface road, the step and the sine signal are the input respectively, the results verify the effectiveness of the union control algorithm.

  • 【网络出版投稿人】 湖南大学
  • 【网络出版年期】2014年 10期
节点文献中: 

本文链接的文献网络图示:

本文的引文网络