【文章內(nèi)容簡介】 器的工作是很重要的。 模糊控制器的設(shè)計包括以下幾項內(nèi)容： 確定模糊控制器的輸入變量和輸出變量； 設(shè)計模糊控制規(guī)則，并計算模糊 控制規(guī)則所決定的模糊關(guān)系，建立模糊控制表； 確立模糊化和非模糊化方法； 合理選擇模糊控制算法的采樣時間。 由于模糊控制器的控制規(guī)則是通過模擬人腦的思維決策方式提出的，所以在選擇模糊控制器的輸入輸出變量時，必須深入研究人在手動控制過程中是如何獲取和輸出信息的。由于人在手動控制過程中，主要是根據(jù)誤差、誤差的變化及誤差的變化的變化來實現(xiàn)控制的，所以模糊控制器的輸入變量也可有三個，即誤差、誤差的變化及誤差的變化的變化，輸出變量一般選擇控制量的變化。 通常將模糊控制器輸入變 量的個數(shù)稱為模糊控制的維數(shù)。由于一般情況下，一維模糊控制器的動態(tài)控制性能并不好，三維模糊控制器的控制規(guī)則過于復(fù)雜，控制算法的實現(xiàn)比較困難，所以，目前被廣泛采用的均為二維模糊控制器，這種控制器以誤差和誤差的變化為輸入變量，以控制量的變化為輸出變量。整個論域即在定義這些模糊子集時應(yīng)注意使論域中任何一點對這些模糊子集的隸屬度的最大值不能太小，否則會在這樣的點附近出現(xiàn)不靈敏區(qū)，以至于造成失控，使模糊控制系統(tǒng)控制性能變壞。 建立模糊控制規(guī)則的基本思想 :當(dāng)誤差大或較大時，選擇控制量以 盡快消除誤差為 主，而當(dāng)誤差較小時，選擇控制量要注意防止超調(diào)，以系統(tǒng)的穩(wěn)定性為主要出發(fā)點。 模糊控制規(guī)則的來源有 3 條途徑 :基于專家經(jīng)驗和實際操作 ,基于模糊模型 ,基于模糊控制的自學(xué)習(xí)。 模糊控制器的控制規(guī)則作為人工手動控制策略的語言描述，它通常用條件語句表示。其主要形式可概括如下： If A then B 9 If A then B else C If A and B then C If A then if B then C If A or B and C or D then E If A then B and if A then C If A then B, C If A then B1 else if A2 then B2 知道上述條件語句之后。以二維模糊控制器為例，假設(shè)條件語句形式為 if E= A then if C= Bj then U=Cij(i=1,2,...,n。j=1,2 ...,m)，式中 Ai Bj Cij 分別定義在誤差、誤差變化和控制量論域 X,Y, Z上的模糊集 。E, C, U 分別代表誤差、誤差變化和控制模糊變量。 一 模糊化方法 由于計算機(jī)采樣輸入的變量均為精確量，所以為便于實現(xiàn)模糊控制算法，須經(jīng)過模糊量化處理變?yōu)槟：俊? 模糊化一般采用如下兩種方法： 將在某區(qū) 間的精確量 x模糊化成這樣的一個模糊子集，它在點 x處隸屬度為 1，除 x 點外其余各點的隸屬均取 0。如所選模糊集合論域為 X={n,n+1,...,0,...,nl,n}，而輸入的基本論域為 [e,e]，輸入精確量為 e。 首先同上算法得到 L,其次查找語言變量賦值表，找出 1 位置上與最大隸屬度所對應(yīng)的語言值所決定的模糊量，該模糊量便為 e 的模糊化量。 二 精確化方法 在模糊控制系統(tǒng)中，由于對建立的模糊控制規(guī)則通過模糊推理決策出的控制變量是一個模糊子集，它不能直接控制被控對象，所以還需要采取合理的方法將其轉(zhuǎn)換為精確 量，以便最好的發(fā)揮出模糊推理結(jié)果的決策效果。 精確化過程的方法很多，主要有 MINMAX 重心法、代數(shù)積 加法 重心法、模糊加權(quán)型推理法、函數(shù)型推理法、加權(quán)函數(shù)型推理法、選擇最大隸屬度法、取中位數(shù)法。 選擇采樣時間是計算機(jī)控制中的構(gòu)性問題，所以模糊控制作為計算機(jī)控制的一種類型，也存在合理的選擇采樣時間的問題。香農(nóng)采樣定理給出了選擇采樣周期的下限 .即 max???T 式中為采樣信號的上限角頻率。 在此范圍內(nèi)，采樣周期越小，就接近連續(xù)控制。但也不能太小，它需要綜合考慮執(zhí)行機(jī)構(gòu)響應(yīng)時間、計 10 算機(jī)控制算法所需時間、計算機(jī)字長、抗干擾性能等多方面因素 The second chapter, the design of fuzzy controller the principle of fuzzy control system Fuzzy control as fuzzy set theory, fuzzy language variable and fuzzy logic reasoning on the basis of a puter numerical control, it has bee the realization of intelligent control is an important and effective especially in the form of fuzzy control and neural work, geic algorithm and the fusion of new disciplines such as chaos theory, is showing its great potential applications. A mon negative feedback control system block diagram in figure 1 By measuring device, controller and controlled object and actuator of the automatic control system, is that people know know the regular feedback control system. Its structure is shown in figure 1. Yet after a longterm research and practice of classical control theory, although for solving the control problem of linear timeinvariant system is very effective. Along with the puter, especially the development and application of microputer based on the type of mu fuzzy quantity, so in order to exert precise control on the controlled, still need to convert their motivation to accurate quantity to u, and then the D/A analog to actuators, for the first step in the control object. Then stop waiting for the second sample, carries on the second step control... This loop is realized with fuzzy control of the controlled object. the basic structure of fuzzy controller The basic structure of fuzzy controller includes knowledge base, fuzzy reasoning, fuzziness of input, output, highprecision four parts. 11 1. The knowledge base Library knowledge base including fuzzy controller parameters and fuzzy control rule base. On the basis of fuzzy control rules based on linguistic variable. Language state variable is the big, , small , such as the fuzzy subset, the fuzzy subset to subordinate function shows that the basic theory of precision value belongs to the fuzzy subset of the domain. Therefore, in order to establish fuzzy control rules, requires the accurate values on the basic theory of domain based on membership function are incorporated into the fuzzy subset, to use the language variable values (large, medium and small, etc.) instead of the accurate values. This process represents the people of observed variables in the control process and control the amount of fuzzy partition. Due to the different variable scope, so the first will be the basic theory of domain respectively in different corresponding relations, mapped to a standardized theory field. Usually, the corresponding relationship between off for quantitative factors. For ease of handling, BiaoZhunLun domain such as bulk chemical separation, and then to fuzzy partition of discourse, define the fuzzy subset, such as NB, PZ, PS, etc. The same fuzzy control rule base, fuzzy partition to the fundamental theory of domain is different, the control effect is also different. Specifically, correspondence, BiaoZhunLun domain, the number of fuzzy subset, and the membership function of fuzzy subset has a great influence on the control effect. These three kinds of parameters and the fuzzy control rules have the same importance, therefore to merge them into fuzzy controller parameter database, together with the fuzzy control rule base of the knowledge base. 2. The blur Convert accurate input into fuzzy quantity F there are two ways: (1) converts gauged BiaoZhunLun fuzzy single point sets on the domain. Gauged by the corresponding relation between x x G into BiaoZhunLun domain on the basic elements, then the elements of the fuzzy single point set F UF (u) = 1 if u = G (x) (2) converts gauged BiaoZhunLun domain of fuzzy subsets. Gauged by the correspon