【作者】 黄治国；

【导师】 张建文；

【作者基本信息】 吉林大学 ， 车辆工程， 2005， 硕士

【摘要】 本文以空气悬架系统为研究对象,结合吉林省科技厅资助项目“汽车电子控制空气悬架的系统开发”,对智能控制方法在空气悬架中的应用以及空气悬架控制系统与虚拟样机技术联合开发的方式进行了一些简单有益的探讨。文中简要介绍了空气弹簧特性和空气悬架对整车性能的影响,并通过台架试验,建立了空气弹簧刚度与电磁阀开关时间的关系。基于神经模糊控制理论及遗传算法,针对1/4 悬架系统,利用Matlab/Simulink 建立了神经模糊控制器,并进行了仿真分析。同时,基于模糊控制理论,在Matlab/Simulink 中建立了1/2 空气悬架控制系统,并利用Real-Time Workshop 将该控制系统导入ADAMS/Car 机械系统虚拟样机模型,然后在ADAMS/Car 中进行联合仿真分析。应用西门子PLC 与微机之间的串口通讯技术,建立了电控空气悬架1/4模型台架试验系统,验证了神经模糊控制算法的有效性。更多还原

【Abstract】 Air suspension is original in the middle of the 19th century .Air suspension is applied mainly to the vibration isolation for mechanical equipments in the early stage. Gradually air suspension has been a kind of standard configuration abroad due to the long time’s researches on air suspension and the perfect performance of air suspension. Electronic control air suspension system is the best air suspension system in the present, which can adjust the parameters of air suspension actively according to the conditions of road surface and the vehicle. Electronic control air suspension system can improve the ride comfort and handling stability of the vehicle and minimize the damage to road surface. With the improvement of the condition of highway, the swell of the consciousness of the road protection, the consummation of the traffic law and regulation and the increased demand for ride comfort, air suspension will be used widely in china. On the other hand, it should be seen that the researches on air suspension in china start early, but china has been behind foreign countries in the application of air suspension for tens years. “The hollowness of technology”is a serious problem which lies in the chinese suspension trade. So we need continual study and research for the accumulation and mastery of the key technology of air suspension. This helps to fill the gap between china and foreign countries in the technology of air suspension and improve the competence and overall quality of chinese automobile technology. The body height control and attitude control can improve the automotive performances to a great extent and it is easy to fulfill the body height control and attitude control in air suspension for its structural convenience, so the design of the structure of control system and control algorithm has been the crux of the improvement of the automotive performances. The researches on the application of intelligent control to air suspension and the co-simulation between virtual prototype and control system have been carried out based on the project “The development of automotive electronic control air suspension system”sponsored by JiLin Science and technology department in this paper. 1. Characteristic analysis of air spring The electronic control air suspension system consists of air compressor、gas storage、electro-magnetic valve、air spring、guided framework、sensor、ECU and so on. Air spring is one of the key parts of the system. This paper briefly present the air spring‘s class、characteristic, such as stiffness characteristic、effective area characteristic and the influence of air spring on the vehicle. What’s more, the relation curve between stiffness of air spring and on/off time of electro-magnetic valve has been obtained through test and the formulas of the relation have been drawn up. Therefore the preparation has been made for the design of fuzzy neural network controller. 2. Fuzzy neural network control of air suspension The published researches on the control algorithm of semi-active suspension and active suspension almost have been concerned with the all branches of modern control theories. Some control algorithms are applied in the suspension control system, but most of them still rest on the academic research stage. Fuzzy neural network control is the focus in the intelligent control and automation field in recent years, which combines fuzzy set theory with neural network. So this paper designs a fuzzy neural network controller by use of Matlab/simulink based on fuzzy neural control theory and genetic algorithm and applies the controller to the quarter-vehicle air suspension system. The learning for fuzzy neural network controller includes two stages: the GA optimization off-line and the modified BP learning on-line. After GA optimization off-line, the fuzzy neural network controller is applied to the fuzzy neural network control system. The learning of modified BP algorithm on-line will be available when the fuzzy neural network control system runs. The simulation is carried out with the random road input. The更多还原