【作者】 张伟谦；

【导师】 潘双夏； 王维锐；

【作者基本信息】 浙江大学 ， 车辆工程， 2010， 硕士

【摘要】 以车辆悬架系统中的半主动减振器和半主动悬架为研究对象,对半主动减振器的结构设计、数学建模和半主动悬架系统控制策略进行了深入研究,主要涉及以下几个方面:(1)提出一种基于遗传算法的磁流变减振器多目标优化设计方法。该方法兼顾对动力性能、响应速度和能耗的要求,能够快速设计出满足性能要求的磁流变减振器。应用该方法实际设计和制造了减振器样机,并通过台架试验对优化设计方法的有效性进行了验证。(2)提出了一种基于修正Backlash滞环的磁流变减振器参数化模型。该模型结构简单,参数意义明确,能准确描述和预测磁流变减振器在不同励磁电流和激振频率下的动特性。(3)研究基于磁流变减振器的半主动悬架系统控制策略。对磁流变减振器的反馈线性化控制、神经网络逆模型控制和Sigmoid逆模型控制进行了对比研究;提出连续型负刚度半主动控制方法,该方法能有效避免开关型负刚度控制中的“颤振”现象,在此基础上提出反馈线性化-连续负刚度、神经网络-连续负刚度和Sigmoid逆模型-连续负刚度三种系统级控制策略,并对其控制效果进行了对比研究。更多还原

【Abstract】 The research objects of this thesis are semi-active damper and semi-active suspension of vehicle suspension system. The structure design of magnetorheological damper, parametric modeling and the control strategies of semi-active suspension system are studied. Main tasks are as follows:(1) A genetic algorithm-based multi-objective optimization of magnetorheological damper design method was presented. The magnetorheological damper could be quickly designed to meet the performance requirements by using the method. The requirements of dynamic performance, response speed and power consumption had been taken into account. A damper prototype was designed and manufactured by using the method. Then, the damper was tested on the bench to vertify the effectiveness of the optimization design method.(2) A modified Backlash hysteresis model was presented. The model was simple in structure. The parameter meaning was clear. The model was able to describe and predict the dynamic characteristics of magnetorheological damper accurately in different currents and frequencies.(3) The semi-active suspension system control strategies based on the magnetorheological damper was studied. The feedback linearization control, neural network inverse model control and sigmoid inverse model control of the magnetorheological damper had been studied contrastively. A continuous negative stiffness semi-active control strategy had been presented, which can avoid the "chatter" phenomenon effectively in the switching of the negative stiffness control. On the basis, three system-level control strategied which called the feedback linearization- continuous negative stiffness, neural networks-continuous negative stiffness and sigmoid inverse model- continuous negative stiffness were presented. The performances of these strategies were studied contrastively.更多还原