【作者】 李伟；

【导师】 杨承；

【作者基本信息】 西南林学院 ， 机械制造及其自动化， 2009， 硕士

【摘要】 电液伺服控制是控制领域中重要的组成部分,具有输出功率大、响应速度快、控制精度高、信号处理灵活、易于实现各种参量的反馈等特点,因而在航空航天、军事、机械、交通、冶金等多个工业领域得到广泛的应用。随着现代工业的发展,控制对象越来越复杂,人们对电液伺服系统的控制品质也提出了更高的要求,应用于电液伺服系统中的一些传统控制策略已经难以满足现代控制的要求,这主要是由电液伺服系统的特性所决定的。电液伺服系统是一种严重的不确定非线性系统,系统工作环境复杂,普遍存在外干扰。在有些控制系统中,控制对象的数学模型很难建立,甚至有的根本无法建模,应用传统的基于模型的控制策略对系统进行设计分析则显得非常困难,而基于知识和不依赖于精确数学模型的智能控制给这类问题的解决带来了新的思路。在人造板热压伺服系统中,由于木材所具有的特殊物理性能,使得受控对象“板材”在热压控制过程中无法建立解析的模型。现在针对此类对象大都采用无模型的PID(Proportion Integral Differential,比例、积分、微分)的控制策略,但是PID参数的确定十分困难,通常需要通过大量的实验分析,才能初步确定可行的参数,而且一旦原材料发生改变,则需要重新通过实验确定控制器参数。针对现有PID控制存在的不足,本文将研究一种基于模糊控制技术的智能控制策略,针对不同的原材料,该策略通过模糊控制技术可以实现控制器的自适应调节。本文的模糊控制器设计如下:采用常见的二维模糊控制器结构,控制器输入系统的偏差e和偏差变化率ec,输出控制量u;依据e的值域,定义模糊控制器的隶属度函数的论域[-5V,+5V];兼顾控制器的运算量与控制精度,在论域中划分7个模糊子集;为便于控制器的软件实现,模糊子集的隶属函数的选定为三角形;由于采用的是二维模糊控制器,所有输入输出精确量都是7个模糊子集的划分,故对应49条模糊规则;推理机使用的是Mamdani推理算法,按照最大隶属度法确定一个最大隶属度y;模糊判决则是通过最大隶属度的平均值法得到模糊控制器的精确输出u。目前智能控制的复杂算法大多是通过程序实现的,虽然一些高端的PLC(Programmable Logic Conctroller,可编程逻辑控制器)中具有模糊控制模块,但其价格相对较高,无疑增加控制系统的成本。随着控制系统的品质不断提高,智能控制的算法也越来越复杂,PLC的内存、运算速度、编程语言都会对复杂算法的实现造成很大的限制。尽管一些PLC的高端产品(如西门子公司的S7-400)具有良好的CPU运算速度,但其价格也相对较高,增加了控制系统的成本。在PC(Personal Computer,个人计算机)中通过灵活多样的编程语言实现控制系统的人机界面及其智能控制算法,则是近些年来智能控制在应用方面的主流趋势。传统的计算机文本编程语言是根据语句和指令的先后顺序决定程序的执行顺序,而美国NI(National Instruments)公司提供的LabVIEW(Laboratory Instrument Engineering Workbench)编程软件是一种图标代替文本行创建应用程序的图形化语言,这种图形化的程序源代码又称G代码。LabVIEW采用并行数据流编程方式,程序框图中节点之间的数据流向决定了程序的执行顺序,它用图标表示函数,用连线表示数据流向。本文正是通过LabVIEW的G代码编程实现了电液位置伺服系统的软件平台——人机界面及其智能控制算法;通过NI公司提供的USB-6008数据采集卡从硬件上连接了计算机控制系统的信号通道——计算机可直接驱动伺服阀放大器。本文首先对电液位置伺服系统建立了数学模型,在MATLAB中仿真了模糊控制策略应用于电液位置伺服系统的控制效果,从理论上证明了此种控制策略相比经典控制的优越性。在系统实现过程中,通过NI公司提供的USB-6008数据采集卡完成了计算机与传感器之间的硬件连接——计算机输入,计算机与伺服阀放大器之间的硬件连接——计算机输出;通过LabVIEW的G代码编程,编制出一套过程可视化的人机界面,在人界面中实现了模糊控制算法及系统参数配置、数据曲线绘制、数据存储等功能。系统仿真和系统运行都表明本文设计的模糊控制策略及其实现方案是可行的。在控制对象相同的条件下,相比传统控制,本文设计的模糊控制器能够有效的减少系统响应时间,实现了控制器的自适应调节;由于本课题使用LabVIEW开发控制系统的应用程序,不仅省略了多种硬件的购买,而且简化了系统的调试过程,有效的节省了应用系统的研发成本,缩短了系统开发时间;使用LabVIEW编程软件编制的应用程序扩展性极强,可生成独立的安装文件在其他没有安装LabVIEW的计算机、触摸屏等多种平台应用。因此,本文为电液位置伺服系统提供了一种可行的智能控制方案。更多还原

【Abstract】 Electro-hydraulic position servo control is an important component in control field, which with features such as great power output, quick response, high precision control, flexible signal processing, many kind of parameters can be easily to feedback, and so on. Therefore it is used in many industry fields such as aeronautics and astronautics, military, mechanism, traffic, metallurgy, and so on.With the development of modern industry, the object which is controlled becomes more and more complex, electro-hydraulic servo system’control quality are put forward to higher requirements, use traditional control strategy on electro-hydraulic servo system are hard to meet the demand of modern control, this is caused by the characteristics of electro-hydraulic servo system. Electro-hydraulic servo system is a kind of serious non-linear system with complex task and environment, which always exist parameter change. In some special control system, it’s hard to build up the mathematical model which is controlled, even some of them is impossible to build up it’s mathematical model, use traditional control strategy on those system would be very hard, but intelligent control which is based on knowledge and not dependent on precise mathematical model show a new way of thinking to solve those problem.In wood-based panel thermal press servo system, because of the special physical characteristics of wood, it’s impossible to build up the wood-based panel’s analytical model when the panel is in thermal press. Now according to these objects which is without model, most strategy to solve them is by PID (Proportion, Integral, Differential), but it’s very hard to confirm PID’s parameters, parameter’s confirm always need many experiments, and if change to another kind of wood, PID’s parameters need re-test to confirm. According to current PID deficiency, this paper would research a kind of intelligent control which is based on fuzzy control, according to different wood this strategy can realize the controller’s adaptive adjust.Fuzzy controller designed in this paper: used common structure of two-dimensional fuzzy controller, controller’s input was deviation e and deviation change rate ec, output was u; according to the range of e, defined the membership function‘s of fuzzy controller domain in [-5V, +5V]; considered the controller’s computation and accuracy, the domain were divided into seven fuzzy sets; In order to facilitate the realization of the controller by software, fuzzy subset membership function selected triangle; because of this paper was designed based on two-dimensional fuzzy controller, all the precise input and output were seven division in fuzzy sets, the fuzzy rules correspond to 49; the Mamdani inference algorithm was used by inference engine , in accordance with the maximum of membership to determine one of the maximum membership degree y; fuzzy decision through the maximum membership degree to achieved the fuzzy controller’s precise output u by average method.Now, most complex algorithms of intelligent control are achieved by program. Although some advanced PLCs(Programmable Logic Controller) contain fuzzy control module,but its cost is relatively high, Undoubtedly increase the cost of control system. With the improvement of control system’s quality,intelligent control algorithms become more an more complex,the algorithm’s realization would be great limited by PLC’s memory, speed of computation, programming languages. Although some advanced PLCs (such as S7-400 provided by SIEMENS) with advanced CPU computing, but its cost is relatively high, too, which also increase the cost of control system. In recent years the application of intelligent control is mainly based on PC (Personal Computer) by which can through flexible programming language to achieve human machine interface and intelligent control algorithm.Traditional computer programming is text languages which are based on sequence of statements to executive program. NI(U.S. National Instruments) Company provide a programming software named LabVIEW (Laboratory Instrument Engineering Workbench), in which uses icon to take place of text to create graphical application program, this kind of graphical program source code is called G-code. LabVIEW’s programming is by parallel data flow, how the data flow indicate means how the order is, it uses icon to express function, uses connection to express how the data flow. This article through LabVIEW’s G-code programming achieved electro-hydraulic position servo system’s software platform—human machine interface and intelligent control algorithm; used the data acquisition card USB-6008 which is provided by NI to connected the computer control system’s signal channel—servo valve amplifier could be directly driven by computer.The mathematical model of electro-hydraulic servo system were built up, used MATLAB simulated the control effect when fuzzy control strategy applied on electro-hydraulic position servo control system, proved theoretically that such a control strategy compared to the classical control showed it’s superiority. During the system’s realization, by using data acquisition card USB-6008 which is offered by NI built up the connection between computer and sensor—computer’s input, connection between computer and the servo valve amplifier—computer’s output; through LabVIEW’s G-code programming realized a set of human machine interface in which with some fuctions such as process visible, fuzzy control algorithm and system parameter’s configuration, data curve drawing, data storage and so on.System simulation and system operation showed that this fuzzy control strategy and its implementation were feasible. When the control object were same, compared to the traditional control strategy, fuzzy controller could reduce the system’s response time effectively and realized the controller’s adaptive adjust; because of this application software were programmed by LabVIEW, not only omitted many hardware’s purchase, but also simplified the process of system’s debugging, the cost of application system development were effectively saved, system’s developing time were shortened; because the application software which was programmed by LabVIEW with good scalability, it can be generated to installation file for computer or touch screen which is not installed LabVIEW. So a kind of feasible intelligent control for electro-hydraulic position servo system was proposed by this paper.更多还原