【作者】 支龙；

【导师】 陈立平；

【作者基本信息】 华中科技大学 ， 机械设计及理论， 2011， 博士

【摘要】 随着汽车技术的不断进步,悬架系统经历了被动悬架、半主动悬架到主动悬架的变迁,设计越来越复杂,对其性能的要求也越来越高。主动悬架系统由于具有优越的综合性能,虽然暂时还未普及,但随着成本的降低以及控制技术、汽车电子技术和新能源技术的发展,相信不久的将来终会成为主流产品。要充分发挥主动悬架硬件的作用,必须有一个好的控制策略软件与其相配合,传统的控制方法(天棚控制、PID控制、滑模控制等)策略成熟,实现简单,对硬件要求低,但在稳定性和鲁棒性以及控制效果等方面仍有很大提高空间。随着智能算法理论的不断完善,智能控制方法已成为现代控制领域研究的热点和主流方向。为了进一步完善汽车主动悬架的控制方法,提高悬架系统的综合性能,本文对汽车主动悬架的智能控制策略进行研究,将在以下几个方面具体开展工作。(1)对汽车悬架系统的简化与建模进行探讨,建立半车四自由度模型和四分之一车二自由度模型,每种模型给出相应的运动微分方程表达式和状态空间表达式。对于路面激励,区分振动和冲击两种不同工况,分别加以研究,并给出两种工况下路面时域激励的具体表达形式。(2)根据所建立的半车四自由度模型,从乘坐舒适性的角度出发,采用模糊控制方法与传统PID控制方法相结合,以使得车身垂直加速度和旋转角加速度最小为控制目标,提出一种新的智能控制策略。用车身的垂直加速度作为PID控制输入变量,车身旋转角速度和角加速度作为模糊控制输入变量,同时以车身和悬架之间允许的相对位移为约束,期望在提高乘坐舒适性的同时,不降低车辆的操控性。(3)同样以半车四自由度模型为研究对象,在考虑悬架弹簧非线性特性的基础上,建立其运动微分方程和状态空间表达式。针对其控制,提出一种模糊PID策略,以车身垂直加速度和速度作为模糊控制器的输入变量,模糊控制器的输出变量是PID控制器的比例、积分和微分增益,最终由PID控制器输出主动控制力。通过仿真方法验证该智能控制算法对降低车身垂直加速度和车身旋转角加速度的控制效果,同时观察车辆操控性能的变化,探讨悬架弹簧的非线性对主动悬架的性能以及控制算法的影响。(4)以四分之一车二自由度模型为研究对象,针对传统PID控制的不足,提出采用分数阶PID控制器对主动悬架进行控制。由于分数阶PID控制比传统PID控制多了微分阶次和积分阶次这两个自由度,因此使得被控系统动态性能的调节更加灵活,但同时也增加了参数整定的难度。参数整定是该控制方法的关键点之一,采用智能算法进行参数整定具有一定的优势,本文将采用遗传优化算法。整个分数阶PID控制器分为两个部分：分数阶PI控制器以车身垂直加速度作为输入,分数阶PD控制器以悬架动挠度作为输入,总的控制力为两控制器输出之和。由于目前计算机软件和硬件控制器都无法直接处理包含分数阶微积分的传递函数,因此在进行仿真计算和实验时,将采用Crone近似法将分数阶传函用一个动态特性足够接近而更易处理的整数阶传函来代替。(5)为验证智能控制策略在主动悬架系统中应用的可靠性及其实际应用效果,将建立主动悬架快速原型实验系统并给出实验方案：基于四分之一车模型,采用半实物仿真搭建主动悬架硬件在回路实验平台,其中控制器使用实物,被控对象用dSPACE模拟。为验证该实验系统及方案是否可行,并验证所提出的智能控制策略的实施效果,将对分数阶PID控制主动悬架和传统PID控制悬架以及被动悬架进行对比实验,对实验结果进行分析。更多还原

【Abstract】 As the automotive technology advances, the suspension system experienced passive suspension, semi-active suspension to active suspension changes. The design of the suspension system is more and more complex, and the performance is required higher and higher. Although active suspension has not been popular for now, it will become mainstream in the future along with the cost reduction and the development of control technology, automobile electronic technology and new energy technology, because of its superior comprehensive performance.A good control strategy has critical influence to full optimal properties of active suspension. The traditional control methods (skyhook control, PID control, sliding mode control, etc.) have many advantages of maturity, easy realization, low hardware requirements, etc., but they remain to be further improved in the stability, robustness and control effect.Intelligent control method has become an intensive research topic and a mainstream of modern control, with the improvement of the intelligent algorithm theory. In order to refine the intelligent control methods for active suspension and increase comprehensive properties of suspension system, this thesis conducts a research on intelligent control strategy for automotive active-suspension and mainly includes following several aspects.(1) Firstly, the simplification and modeling of automotive suspension system are discussed. Mathematics models of the half car 4-degree-of-freedom and the quarter car 2-degree-of-freedom are established, respectively. The differential equations of motion and state-space expressions of each model are given. Road excitation is divided into two different conditions:vibration and impact, which be studied respectively. The specific expressions of time domain road input in different conditions are present.(2) An active suspension model based on PID-FUZZY control is built. The half car model has four degrees of freedom, subject to irregular excitation from a road surface. The control is the sum of PID and fuzzy-logic control. PID control uses the vertical acceleration of vehicle as the input source, fuzzy-logic control uses the rotary acceleration and velocity as the input source. The simulation results show that the active suspension is effective in vibration isolation of vehicle body.(3) An active suspension system for vehicles using fuzzy-pid control is proposed. The half-car model treated here is described by a nonlinear system with four degrees of freedom, subject to excitation from a road surface. The active control is fuzzy-pid control, and the control is determined from the view-point of ride comfort, minimizing the vertical and rotary acceleration of vehicle body. In the derivation of fuzzy control rules, vertical velocity and acceleration of the vehicle body are denoted as the input variables. The simulation results indicate that the proposed active suspension performs effectively when compared with other forms of control.(4) An active suspension for the quarter car 2-degree-of-freedom model using a fractional PIλDμcontroller is present. The active control is expressed as the summation of PIλand PDμthat choose vertical acceleration of the body and suspension deflection as input variable, respectively. The parameters of the controller are optimized using the genetic algorithm (GA). The simulation results indicate that the fractional PIλDμcontroller for suspension system is effective in the vibration isolation of the vehicle body and has better performance than classical PID controller.(5) In order to verify the control effect of the control strategy proposed in this paper, the hardware in the loop simulation test platform based on the dSPACE real time system is built. The road input and states of the suspension system will be simulated in high precision by using the HIL test platform, which has advantages of simple structure, short time for ready and low test cost. The effects of the intelligent control strategy and fractional PIλDμcontrol algorithm are verified by the hardware in the loop test platform, the performances of suspension system are improved evidently. The HIL test also provided important reference for the further research of active suspension control system.更多还原