【作者】 魏志毅；

【导师】 孔祥东；

【作者基本信息】 燕山大学 ， 机械电子工程， 2007， 硕士

【摘要】 电液比例系统是工业生产、科学研究中常见和最基本的电液系统之一,广泛应用于社会生活的各个领域。由于电液比例系统是结构复杂的机、电、液综合系统,普遍存在非线性、时变性和不确定性,且系统精确的数学模型不易建立,应用传统的PID控制理论不易解决电液比例系统中存在的一些问题,因此迫切需要寻找新的控制策略。随着科学技术的进步,控制理论也在不断地发展。模糊控制和神经网络控制都是正在兴起的控制技术,本文将模糊控制和神经网络相结合用于电液比例压力控制系统,通过系统地分析模糊控制和神经网络控制系统的结构、算法及系统应用等问题,集模糊控制和神经网络控制技术之长,构建了基于神经网络的模糊控制技术,进行了理论研究与算法实现。以神经网络的层和节点分别对应模糊系统的各个部分,将模糊控制的规则隐含地分布在整个网络中,用神经网络实现模糊推理。这种控制策略充分利用了模糊控制既符合人们对过程作用的直观描述和思维逻辑的优点,又解决了神经网络参数没有确切意义而难以理解的问题。本文以材料试验机为研究平台,通过系统辨识的方法确定了材料试验机电液比例压力控制系统的数学模型,利用MATLAB/Simulink仿真工具分别构建了PID控制器和模糊神经网络控制器,并对其进行了动态建模仿真分析,最后用半实物仿真工具dSPACE实现了电液比例压力控制系统的控制策略的实验研究及验证。结果表明:与PID控制系统相比,模糊神经网络控制具有鲁棒性好,超调量小,自适应能力强等优点。更多还原

【Abstract】 Electro-hydraulic proportional system is one of the ordinary and essential electro–hydraulic systems in the industry and scientific research. Electro-hydraulic proportional system is applied to many areas of the society. Electro-hydraulic proportional system is a typical system which is coupled with mechanical technology, electric technology and hydraulic technology. Non-linearity, non-confirmation, outside and across disturbances are existed prevalently in the electro-hydraulic proportional system and the accurate mathematic model of the system is established difficultly. The traditional PID control theory can’t solve the control problem on the electro-hydraulic proportional system, so we must find a new control tactic immediately.With the progress of science and technique, the theory of control is developing too. Fuzzy control and neural network control is new control technique. In this paper, the combination of fuzzy control and neural network control is applied to the electro-hydraulic proportional system. The structure、arithmetic and application of fuzzy control and neural network control are analyzed systemically. Used the merits of fuzzy control and neural network control, the technology of fuzzy control with neural network-based is created. The layer and node of neural network are corresponding to every part of the fuzzy system, and the rules of fuzzy control are distributed in the whole network impliedly, so that fuzzy reasoning is implemented by neural network. Fuzzy control is sufficiently used in this control strategy, it is according with the intuitionistic description and logistic thought. At the same time, the method solves the problem that the parameters of neural network is difficult to understand because the parameters don’t have precise sense.This paper is based on the material testing machine. The mathematic model of the electro-hydraulic proportional system of pressure on the machine is confirmed by system identification. Used the software MATLAB/Simulink, the PID controller and the fuzzy neural network controller were created, and modeled and simulated for them. Then, through the semi physical simulation software of dSPACE, the control strategy of theelectro-hydraulic proportional system of pressure was realized in experiment and proofed. The results indicate: compared with the PID control, the fuzzy neural network control has more merits, such as better robustness, letter overshoot and more powerful self-adapting capability, etc.更多还原