模糊控制畢業(yè)論文模糊控制的理論與發(fā)展概述

1、<p>　　題目:模糊控制的理論與發(fā)展概述</p><p>　　摘要 模糊控制理論是以模糊數(shù)學(xué)為基礎(chǔ),用語言規(guī)則表示方法和先進(jìn)的計(jì)算機(jī)技術(shù),由模糊推理進(jìn)行決策的一種高級(jí)控制策。模糊控制作為以模糊集合論、模糊語言變量及模糊邏輯推理為基礎(chǔ)的一種計(jì)算機(jī)數(shù)字控制，它已成為目前實(shí)現(xiàn)智能控制的一種重要而又有效的形式尤其是模糊控制和神經(jīng)網(wǎng)絡(luò)、遺傳算法及混沌理論等新學(xué)科的融合，正在顯示出其巨大的應(yīng)用潛力。實(shí)質(zhì)上模糊控制

2、是一種非線性控制，從屬于智能控制的范疇。模糊控制的一大特點(diǎn)是既具有系統(tǒng)化的理論，又有著大量實(shí)際應(yīng)用背景。</p><p>　　本文簡單介紹了模糊控制的概念及應(yīng)用，詳細(xì)介紹了模糊控制器的設(shè)計(jì)，其中包含模糊控制系統(tǒng)的原理、模糊控制器的分類及其設(shè)計(jì)元素。</p><p>　　關(guān)鍵詞：模糊控制；模糊控制器；現(xiàn)狀及展望</p><p>　　Abstract Fuzzy co

3、ntrol theory is based on fuzzy mathematics, using language rule representation and advanced computer technology, it is a high-level control strategy which can make decision by the fuzzy reasoning. Fuzzy control is a comp

4、uter numerical contro which based fuzzy set theory, fuzzy linguistic variables and fuzzy logic, it has become the effective form of intelligent control especially in the form of fuzzy control and neural networks, genetic

5、 algorithms and chaos theory and other new</p><p>　　This article introduces simply the concept and application of fuzzy control and introduces detailly the design of the fuzzy controller. It contains the pr

6、inciples of fuzzy control system, the classification of fuzzy controller and its design elements.</p><p>　　Key words: Fuzzy Control; Fuzzy Controller; Status and Prospects.</p><p><b>

7、　　引言</b></p><p>　　傳統(tǒng)的常規(guī)PID控制方式是根據(jù)被控制對(duì)象的數(shù)學(xué)模型建立，雖然它的控制精度可以很高，但對(duì)于多變量且具有強(qiáng)耦合性的時(shí)變系統(tǒng)表現(xiàn)出很大的誤差。比例調(diào)節(jié)是根據(jù)被調(diào)量和設(shè)定值之間的差值來變化的，也就是說比例控制中余差不可避免。積分調(diào)節(jié)最終實(shí)現(xiàn)無余差調(diào)節(jié)，但是超調(diào)比較大。模糊控制是建立在人工經(jīng)驗(yàn)基礎(chǔ)之上的,它能將熟練操作員的實(shí)踐經(jīng)驗(yàn)加以總結(jié)和描述,并用語言表達(dá)出來,得到定性的

8、、精確的控制規(guī)則,不需要被控對(duì)象的數(shù)學(xué)模型。并且模糊控制易于被人們接受,構(gòu)造容易,適應(yīng)性好。</p><p>　　The introduction</p><p>　　Traditional way of conventional PID control was established according to the mathematical model of controlled ob

9、ject, although it can be very high control precision, but for the multi-variable</p><p>　　and time-varying systems with strong coupling showed great error. Proportional control is based on the difference in

10、value between set value and quantity of the modulated to change, that is to say, proportional control of residual is inevitable. Integral regulation achieve everything in a glance at poor regulation, but the overshoot is

11、 bigger. Fuzzy control is based on the artificial experience, it can skilled operator's practical experience summarized and described, and the language expression,</p><p>　　第一章 模糊控制概述</p><p>

12、;　　1.1模糊控制的概念及應(yīng)用</p><p>　　“模糊”是人類感知萬物,獲取知識(shí),思維推理,決策實(shí)施的重要特征?！澳：北取扒逦彼鶕碛械男畔⑷萘扛?內(nèi)涵更豐富,更符合客觀世界。模糊邏輯控制(Fuzzy Logic Control)簡稱模糊控制(Fuzzy Control)，是以模糊集合論、模糊語言變量和模糊邏輯推理為基礎(chǔ)的一種計(jì)算機(jī)數(shù)字控制技術(shù)。模糊控制理論是由美國著名的學(xué)者加利福尼亞大學(xué)教授Zadeh

13、·L·A于1965年首先提出,它是以模糊數(shù)學(xué)為基礎(chǔ),用語言規(guī)則表示方法和先進(jìn)的計(jì)算機(jī)技術(shù),由模糊推理進(jìn)行決策的一種高級(jí)控制策。在1968～1973年期間Zadeh·L·A先后提出語言變量、模糊條件語句和模糊算法等概念和方法，使得某些以往只能用自然語言的條件語句形式描述的手動(dòng)控制規(guī)則可采用模糊條件語句形式來描述，從而使這些規(guī)則成為在計(jì)算機(jī)上可以實(shí)現(xiàn)的算法。1974年,英國倫敦大學(xué)教授Mamdani&

14、#183;E·H研制成功第一個(gè)模糊控制器, 并把它應(yīng)用于鍋爐和蒸汽機(jī)的控制，在實(shí)驗(yàn)室獲得成功。這一開拓性的工作標(biāo)志著模糊控制論的誕生并充分展示了模糊技術(shù)的應(yīng)用前景。</p><p>　　模糊控制實(shí)質(zhì)上是一種非線性控制，從屬于智能控制的范疇。模糊控制的一大特點(diǎn)是既具有系統(tǒng)化的理論，又有著大量實(shí)際應(yīng)用背景。模糊控制的發(fā)展最初在西方遇到了較大的阻力；然而在東方尤其是在日本，卻得到了迅速而廣泛的推廣應(yīng)用。其典型

15、應(yīng)用的例子涉及生產(chǎn)和生活的許多方面， 以下為模糊控制在工業(yè)和生活方面的一些應(yīng)用實(shí)例：</p><p>　　1.2模糊控制的優(yōu)點(diǎn) </p><p>　　1簡化系統(tǒng)設(shè)計(jì)的復(fù)雜性，特別適用于非線性、時(shí)變、模型不完全的系統(tǒng)上。 </p><p>　　2利用控制法則來描述系統(tǒng)變量間的關(guān)系。 </p><p>　　3不用數(shù)值而用語言式的模糊變量來描述系統(tǒng)

16、，模糊控制器不必對(duì)被控制對(duì)象建立完整的數(shù)學(xué)模式。 </p><p>　　4模糊控制器是一語言控制器，使得操作人員易于使用自然語言進(jìn)行人機(jī)對(duì)話。 </p><p>　　5模糊控制器是一種容易控制、掌握的較理想的非線性控制器，并且抗干擾能力強(qiáng)，響應(yīng)速度快，并對(duì)系統(tǒng)參數(shù)的變化有較強(qiáng)的魯棒性和較佳的容錯(cuò)性。 </p><p>　　6從屬于智能控制的范疇。該系統(tǒng)尤其適于非線性

17、，時(shí)變，滯后系統(tǒng)的控制。</p><p>　　1.3模糊控制的缺點(diǎn) </p><p>　　1模糊控制的設(shè)計(jì)尚缺乏系統(tǒng)性，這對(duì)復(fù)雜系統(tǒng)的控制是難以奏效的。所以如何建立一套系統(tǒng)的模糊控制理論，以解決模糊控制的機(jī)理、穩(wěn)定性分析、系統(tǒng)化設(shè)計(jì)方法等一系列問題； </p><p>　　2 如何獲得模糊規(guī)則及隸屬函數(shù)即系統(tǒng)的設(shè)計(jì)辦法，這在目前完全憑經(jīng)驗(yàn)進(jìn)行； </p>

18、<p>　　3 信息簡單的模糊處理將導(dǎo)致系統(tǒng)的控制精度降低和動(dòng)態(tài)品質(zhì)變差。若要提高精度則必然增加量化級(jí)數(shù)，從而導(dǎo)致規(guī)則搜索范圍擴(kuò)大，降低決策速度，甚至不能實(shí)時(shí)控制； </p><p>　　4.如何保證模糊控制系統(tǒng)的穩(wěn)定性即如何解決模糊控制中關(guān)于穩(wěn)定性和魯棒性問題還有待完善。</p><p>　　The first chapter is summary of fuzzy con

19、trol</p><p>　　the concept and application of fuzzy control</p><p>　　"Fuzzy" human perception is everything, to acquire knowledge, thinking, reasoning, decision-making of important feat

20、ures. "Fuzzy" than "clear" have the information capacity of a larger, more abundant connotation, more in line with the objective world. Fuzzy Logic Control (Fuzzy Logic Control) referred to as "F

21、uzzy Control (Fuzzy Control), based on the Fuzzy set theory, Fuzzy language variable and Fuzzy Logic reasoning is the basis of a computer numerical Control technology. Fuzzy control theory </p><p>　　Fuzzy co

22、ntrol is essentially a kind of nonlinear control, from belongs to the category of intelligent control. Fuzzy control is one of the biggest characteristic is both a systematic theory, and with a large number of practical

23、application background. The development of fuzzy control is first encountered in the west the larger resistance; In the east, especially in Japan, however, has obtained the rapid and extensive popularization and applicat

24、ion. Its typical application example involves many a</p><p>　　the advantages of fuzzy control</p><p>　　1 to simplify the complexity of system design, especially for nonlinear, time-varying and m

25、odel incomplete system.</p><p>　　2 control law is used to describe the relationship between the system variables.</p><p>　　3 language instead of numerical type of fuzzy variables to describe the

26、 system, the fuzzy controller don't need to establish a comprehensive mathematical model of controlled object.</p><p>　　4 controller, fuzzy controller is a language that operators easy to use natural lan

27、guage for the man-machine dialogue.</p><p>　　5 fuzzy controller is a kind of easy to control and mastery of the ideal nonlinear controller, and strong anti-jamming capability, fast response speed, and the ch

28、ange of system parameters has strong robustness and better fault tolerance.</p><p>　　6 from belongs to the category of intelligent control. This system is especially suitable for nonlinear, time varying and

29、lag control system.</p><p>　　1.3 the disadvantage of fuzzy control</p><p>　　1 the design of the fuzzy control is still lack of systematic, the control of complex systems is difficult to work. So

30、 how to establish a system of fuzzy control theory, in order to solve fuzzy control mechanism, stability analysis, systematic design method for a series of problems;</p><p>　　2 how to obtain fuzzy rules and

31、membership functions, system design, complete with experience for this in the present;</p><p>　　3 simple fuzzy information processing will reduce control precision of the system and the dynamic quality becom

32、es poor. If you want to improve the accuracy of inevitable to increase the quantitative series, leading to rule search scope, reduce the decision-making speed, can't even real-time control;</p><p>　　4. H

33、ow to ensure the stability of the fuzzy control system is how to solve the fuzzy control on the stability and robustness problems remains to be perfect.</p><p>　　第二章 模糊控制器的設(shè)計(jì)</p><p>　　2.1 模糊控制系統(tǒng)

34、的原理</p><p>　　模糊控制作為以模糊集合論、模糊語言變量及模糊邏輯推理為基礎(chǔ)的一種計(jì)算機(jī)數(shù)字控制，它已成為目前實(shí)現(xiàn)智能控制的一種重要而又有效的形式尤其是模糊控制和神經(jīng)網(wǎng)絡(luò)、遺傳算法及混沌理論等新學(xué)科的融合，正在顯示出其巨大的應(yīng)用潛力。</p><p>　　圖1 常見負(fù)反饋控制系統(tǒng)方框圖</p><p>　　由測(cè)量裝置、控制器、被控對(duì)象及執(zhí)行機(jī)構(gòu)組成的自動(dòng)控

35、制系統(tǒng)，就是人們所悉知的常規(guī)負(fù)反饋控制系統(tǒng)。其結(jié)構(gòu)如圖1所示。然而經(jīng)過人們長期研究和實(shí)踐形成的經(jīng)典控制理論，雖然對(duì)于解決線性定常系統(tǒng)的控制問題非常有效。隨著計(jì)算機(jī)尤其是微機(jī)的發(fā)展和應(yīng)用，基于由于式中μ模糊量，所以為了對(duì)被控對(duì)象施加精確的控制，還需要將其清晰化轉(zhuǎn)換為精確量u，然后經(jīng)D/A得模擬量送給執(zhí)行機(jī)構(gòu)，對(duì)被對(duì)象進(jìn)行第一步控制。然后中斷等待第二次采樣，進(jìn)行第二步控制...這樣循環(huán)下去就實(shí)現(xiàn)了對(duì)被控對(duì)象的模糊控制。</p>

36、<p>　　2.2模糊控制器的基本結(jié)構(gòu)</p><p>　　模糊控制器的基本結(jié)構(gòu)包括知識(shí)庫、模糊推理、輸入量模糊化、輸出量精確化四部分。</p><p><b>　　1.知識(shí)庫 </b></p><p>　　知識(shí)庫包括模糊控制器參數(shù)庫和模糊控制規(guī)則庫。模糊控制規(guī)則建立在語言變量的基礎(chǔ)上。語言變量取值為“大”、“中”、“小”等這樣的模

37、糊子集,各模糊子集以隸屬函數(shù)表明基本論域上的精確值屬于該模糊子集的程度。因此,為建立模糊控制規(guī)則,需要將基本論域上的精確值依據(jù)隸屬函數(shù)歸并到各模糊子集中,從而用語言變量值(大、中、小等)代替精確值。這個(gè)過程代表了人在控制過程中對(duì)觀察到的變量和控制量的模糊劃分。由于各變量取值范圍各異,故首先將各基本論域分別以不同的對(duì)應(yīng)關(guān)系,映射到一個(gè)標(biāo)準(zhǔn)化論域上。通常,對(duì)應(yīng)關(guān)系取為量化因子。為便于處理,將標(biāo)準(zhǔn)論域等分離散化,然后對(duì)論域進(jìn)行模糊劃分,定義模

38、糊子集,如NB、PZ、PS等。 </p><p>　　同一個(gè)模糊控制規(guī)則庫,對(duì)基本論域的模糊劃分不同,控制效果也不同。具體來說,對(duì)應(yīng)關(guān)系、標(biāo)準(zhǔn)論域、模糊子集數(shù)以及各模糊子集的隸屬函數(shù)都對(duì)控制效果有很大影響。這3類參數(shù)與模糊控制規(guī)則具有同樣的重要性,因此把它們歸并為模糊控制器的參數(shù)庫,與模糊控制規(guī)則庫共同組成知識(shí)庫。 </p><p><b>　　2.模糊化 </b>&

39、lt;/p><p>　　將精確的輸入量轉(zhuǎn)化為模糊量F有兩種方法: </p><p>　　(1)將精確量轉(zhuǎn)換為標(biāo)準(zhǔn)論域上的模糊單點(diǎn)集。精確量x經(jīng)對(duì)應(yīng)關(guān)系G轉(zhuǎn)換為標(biāo)準(zhǔn)論域x上的基本元素,則該元素的模糊單點(diǎn)集F為 </p><p>　　uF(u)=1 if u=G(x) </p><p>　　(2)將精確量轉(zhuǎn)換為標(biāo)準(zhǔn)論域上的模糊子集。 </p&g

40、t;<p>　　精確量經(jīng)對(duì)應(yīng)關(guān)系轉(zhuǎn)換為標(biāo)準(zhǔn)論域上的基本元素,在該元素上具有最大隸屬度的模糊子集,即為該精確量對(duì)應(yīng)的模糊子集。 </p><p><b>　　3.模糊推理 </b></p><p>　　最基本的模糊推理形式為: </p><p>　　前提1　IF A THEN B </p><p>　　前提2

41、　IF A′ </p><p>　　結(jié)論　THEN B′ </p><p>　　其中,A、A′為論域U上的模糊子集,B、B′為論域V上的模糊子集。前提1稱為模糊蘊(yùn)涵關(guān)系,記為A→B。在實(shí)際應(yīng)用中,一般先針對(duì)各條規(guī)則進(jìn)行推理,然后將各個(gè)推理結(jié)果總合而得到最終推理結(jié)果。 </p><p><b>　　4.精確化 </b></p>&l

42、t;p>　　推理得到的模糊子集要轉(zhuǎn)換為精確值,以得到最終控制量輸出y。目前常用兩種精確化方法: </p><p>　　(1)最大隸屬度法。在推理得到的模糊子集中,選取隸屬度最大的標(biāo)準(zhǔn)論域元素的平均值作為精確化結(jié)果。 </p><p>　　(2)重心法。將推理得到的模糊子集的隸屬函數(shù)與橫坐標(biāo)所圍面積的重心所對(duì)應(yīng)的標(biāo)準(zhǔn)論域元素作為精確化結(jié)果。在得到推理結(jié)果精確值之后,還應(yīng)按對(duì)應(yīng)關(guān)系,得到

43、最終控制量輸出y。</p><p>　　2.3模糊控制器的分類</p><p><b>　　模糊控制的類型有:</b></p><p>　　(1)基本模糊控制器：一旦模糊控制表確定之后,控制規(guī)則就固定不變了;</p><p>　　(2)自適應(yīng)模糊控制器：在運(yùn)行中自動(dòng)修改、完善和調(diào)整規(guī)則,使被控過程的控制效果不斷提高,達(dá)到

44、預(yù)期的效果；</p><p>　　(3)智能模糊控制器：它把人、人工智能和神經(jīng)網(wǎng)絡(luò)三者聯(lián)系起來,實(shí)現(xiàn)綜合信息處理,使系統(tǒng)既具有靈活的推理機(jī)制、啟發(fā)性知識(shí)與產(chǎn)生式規(guī)則表示,又具有多種層次、多種類型的控制規(guī)律選擇。</p><p>　　2.4模糊控制器的設(shè)計(jì)</p><p>　　模糊控制器在模糊自動(dòng)控制系統(tǒng)中具有舉足輕重的作用，因此在模糊控制系統(tǒng)中，設(shè)計(jì)和調(diào)整模糊控制器

45、的工作是很重要的。</p><p>　　模糊控制器的設(shè)計(jì)包括以下幾項(xiàng)內(nèi)容：</p><p>　　1、確定模糊控制器的輸入變量和輸出變量；</p><p>　　2、設(shè)計(jì)模糊控制規(guī)則，并計(jì)算模糊控制規(guī)則所決定的模糊關(guān)系，建立模糊控制表；</p><p>　　3、確立模糊化和非模糊化方法；</p><p>　　4、合理選擇模

46、糊控制算法的采樣時(shí)間。</p><p>　　2.4.1模糊控制器的輸入輸出變量</p><p>　　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(shí)，必須深入研究人在手動(dòng)控制過程中是如何獲取和輸出信息的。由于人在手動(dòng)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(shí)現(xiàn)控制的，所以模糊控制器的輸入變量也可有三個(gè)，即誤差、誤差的變化及誤

47、差的變化的變化，輸出變量一般選擇控制量的變化。</p><p>　　通常將模糊控制器輸入變量的個(gè)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(dòng)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(shí)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(gè)論域即在定義這些模糊子集時(shí)應(yīng)注意使論域中任何一點(diǎn)對(duì)這些模糊子集的隸屬度的最大

48、值不能太小，否則會(huì)在這樣的點(diǎn)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。</p><p>　　2.4.2建立模糊控制器的控制規(guī)則</p><p>　　建立模糊控制規(guī)則的基本思想:當(dāng)誤差大或較大時(shí)，選擇控制量以盡快消除誤差為</p><p>　　主，而當(dāng)誤差較小時(shí)，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(diǎn)。</p><

49、p>　　模糊控制規(guī)則的來源有3條途徑:基于專家經(jīng)驗(yàn)和實(shí)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(dòng)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：</p><p>　　If A then B</p><p>　　If A then B else C</p><p>　　If A and B then C</

50、p><p>　　If A then if B then C</p><p>　　If A or B and C or D then E</p><p>　　If A then B and if A then C</p><p>　　If A then B, C</p><p>　　If A then B1 else if

51、A2 then B2</p><p>　　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。</p><p>　　2.4.

52、3確立模糊化和精確化化方法</p><p><b>　　一 模糊化方法</b></p><p>　　由于計(jì)算機(jī)采樣輸入的變量均為精確量，所以為便于實(shí)現(xiàn)模糊控制算法，須經(jīng)過模糊量化處理變?yōu)槟：俊?lt;/p><p>　　模糊化一般采用如下兩種方法：</p><p>　　1、將在某區(qū)間的精確量x模糊化成這樣的一個(gè)模糊子集，它在

53、點(diǎn)x處隸屬度為1，除x點(diǎn)外其余各點(diǎn)的隸屬均取0。如所選模糊集合論域?yàn)閄={-n,-n+1,...,0,...,n-l,n}，而輸入的基本論域?yàn)閇-e,e]，輸入精確量為e。</p><p>　　2、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(duì)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。</p><p><b>　　二 精確化方法</b&g

54、t;</p><p>　　在模糊控制系統(tǒng)中，由于對(duì)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(gè)模糊子集，它不能直接控制被控對(duì)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。</p><p>　　精確化過程的方法很多，主要有MIN-MAX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。

55、</p><p>　　2.4.4采樣時(shí)間的選擇</p><p>　　選擇采樣時(shí)間是計(jì)算機(jī)控制中的構(gòu)性問題，所以模糊控制作為計(jì)算機(jī)控制的一種類型，也存在合理的選擇采樣時(shí)間的問題。香農(nóng)采樣定理給出了選擇采樣周期的下限.即</p><p>　　式中為采樣信號(hào)的上限角頻率。</p><p>　　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，

56、它需要綜合考慮執(zhí)行機(jī)構(gòu)響應(yīng)時(shí)間、計(jì)算機(jī)控制算法所需時(shí)間、計(jì)算機(jī)字長、抗干擾性能等多方面因素</p><p>　　The second chapter, the design of fuzzy controller</p><p>　　2.1 the principle of fuzzy control system</p><p>　　Fuzzy control a

57、s fuzzy set theory, fuzzy language variable and fuzzy logic reasoning on the basis of a computer numerical control, it has become the realization of intelligent control is an important and effective especially in the for

58、m of fuzzy control and neural network, genetic algorithm and the fusion of new disciplines such as chaos theory, is showing its great potential applications.</p><p>　　A common negative feedback control syste

59、m block diagram in figure 1</p><p>　　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 stru

60、cture is shown in figure 1. Yet after a long-term research and practice of classical control theory, although for solving the control problem of linear time-invariant system is very effective. Along with the computer, es

61、pecially the development and application of microcomputer based on the type of mu fuzzy quantity, so in order </p><p>　　2.2 the basic structure of fuzzy controller</p><p>　　The basic structure o

62、f fuzzy controller includes knowledge base, fuzzy reasoning, fuzziness of input, output, high-precision four parts.</p><p>　　1. The knowledge base</p><p>　　Library knowledge base including fuzzy

63、 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 su

64、bset, 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

65、on the basic theory of domain based on membership fu</p><p>　　The same fuzzy control rule base, fuzzy partition to the fundamental theory of domain is different, the control effect is also different. Specifi

66、cally, 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

67、have the same importance, therefore to merge them into fuzzy controller parameter database, together with the fuzzy control rule base o</p><p>　　2. The blur</p><p>　　Convert accurate input into

68、fuzzy quantity F there are two ways:</p><p>　　(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 el

69、ements, then the elements of the fuzzy single point set F</p><p>　　UF (u) = 1 if u = G (x)</p><p>　　(2) converts gauged BiaoZhunLun domain of fuzzy subsets.</p><p>　　Gauged by the c

70、orresponding relationship into BiaoZhunLun domain on the basic elements, on the element has the maximum membership degree of fuzzy subsets, namely for the precise amount corresponding fuzzy subset.</p><p>　　

71、3. The fuzzy inference</p><p>　　The most basic form of fuzzy reasoning is:</p><p>　　1 IF A THEN B</p><p>　　Premise 2 IF A '</p><p>　　Conclusion THEN B '</p&g

72、t;<p>　　Among them, A, A 'for fuzzy subset on the theory of domain U, B, B' for the theory of fuzzy subset V on the domain. Premise 1 is called the fuzzy implication relations, to A and B. In practice, the

73、 general rules of first in view of the individual reasoning, then the reasoning result sum and eventually reasoning results are obtained.</p><p>　　4. Accurate</p><p>　　Reasoning of fuzzy subset

74、to convert accurate value, to get the final control output y. Two accurate methods commonly used at present:</p><p>　　(1) the maximum membership degree method. In the reasoning of fuzzy subset, the selection

75、 membership degree of the largest BiaoZhunLun domain element as the average of the accurate results.</p><p>　　(2) the gravity method. Will get the membership function of fuzzy subset reasoning abscissa and t

76、he area around the center of gravity of the corresponding BiaoZhunLun domain elements as accurate results. After the reasoning results precision value, still should according to corresponding relation, get the final cont

77、rol output y.</p><p>　　2.3 the classification of the fuzzy controller</p><p>　　The type of fuzzy control are:</p><p>　　(1) the basic fuzzy controller: after fuzzy control table was

78、determined, control rules are fixed;</p><p>　　(2) the adaptive fuzzy controller: in the operation of the automatic change, improve and adjust the rules and make the control effect of increasing the charged p

79、rocess and achieve the desired effect;</p><p>　　(3) the intelligent fuzzy controller, it brings people, artificial intelligence and neural network, and realize the integrated information processing, make the

80、 system both with flexible inference mechanism, the heuristic knowledge and production rule says, but also has many layers, choice of multiple types of control law.</p><p><b>　　窗體頂端</b></p>

81、<p>　　2.4 the design of fuzzy controller</p><p>　　Fuzzy controller plays an important role in fuzzy automatic control system, so the fuzzy control system, the design and adjustment of fuzzy controller

82、 is very important.</p><p>　　The design of fuzzy controller includes the following content:</p><p>　　1, to determine the fuzzy controller input variables and output variables;</p><p&g

83、t;　　2, the design of fuzzy control rules, and calculate the fuzzy control rules determined by fuzzy relation, fuzzy control table;</p><p>　　3, establish fuzzy and the fuzzy method;</p><p>　　4, t

84、he rational selection of sampling time fuzzy control algorithm.</p><p>　　2.4.1 fuzzy controller input and output variables</p><p>　　Because the control rules of fuzzy controller is by simulating

85、 the human brain thinking decision-making mode is put forward, so when choosing input and output variables of the fuzzy controller, must further study in the process of manual control is how to obtain and output informat

86、ion. Because people in the process of manual control, mainly according to the change of the error, error change and error changes to realize the control, so the input variable of fuzzy controller also can have three, i&l

87、t;/p><p>　　Will the number of input variable of fuzzy controller usually referred to as the dimensions of the fuzzy control. Because in general, a one-dimensional dynamic control performance of fuzzy controller

88、 is not good, the three dimensional fuzzy controller control rules are too complex, the realization of the control algorithm is more difficult, so the current widely used are based on two-dimensional fuzzy controller, th

89、is controller with error and error change as input variables, to control the chan</p><p>　　2.4.2 establish control rules of fuzzy controller</p><p>　　Set up the basic idea of fuzzy control rules

90、: when the error is big or large, choose control quantity in order to eliminate the error is as soon as possible</p><p>　　Lord, when the error is small, select control should pay attention to prevent oversho

91、ot, system stability as the main starting point.</p><p>　　The source of the fuzzy control rules has three ways: based on the expert experience and the actual operation, based on the fuzzy model, based on the

92、 fuzzy control of self learning. Control rules of fuzzy controller as a manual control strategy of language description, it is usually expressed in conditional statements. Its main forms can be summarized as follows:<

93、/p><p>　　If A then B</p><p>　　If A then B else C</p><p>　　If A and B then C</p><p>　　If A then if B then C</p><p>　　If A or B and C or D then E</p><

94、;p>　　If A then B and if A then C</p><p>　　If A then B, C</p><p>　　If A then B1 else if A2 then B2</p><p>　　After know the conditional statement. In two-dimensional fuzzy controll

95、er, for example, suppose that conditional statements form as if E = A then if C = Bj then U = Cij (I = 1, 2,..., n; j = 1, 2,..., m), type of Ai Bj Cij respectively defined in the theory of error, error change and contro

96、l the amount of X, Y, Z in fuzzy sets; E, C, U represent the error, error change and control of fuzzy variables.</p><p>　　2.4.3 establish fuzzy and accurate method</p><p>　　A fuzzy method</p&

97、gt;<p>　　The computer sampling the input variables are based on accurate quantity, so the fuzzy control algorithm for ease of implementation, must be made via fuzzy quantization process into fuzzy quantity.</p&

98、gt;<p>　　Blur generally USES the following two methods:</p><p>　　1, will be in a certain range of gauged x fuzzy into such a fuzzy subset, it at point x membership for 1, in addition to the x each poi

99、nt in the rest of the membership are 0. Domains such as the selected fuzzy set theory to X = {- n - n + 1,... , 0,... , n - l, n}, and input the basic theory of domain for [- e, e], accurate quantity input is e.</p>

100、;<p>　　2, first ditto algorithm to get the L, second language variable assignment table lookup, find out 1 position with the maximum membership degree is determined by the linguistic values of the corresponding fu

101、zzy quantity, the fuzzy measure for e the blur.</p><p>　　Accurate method</p><p>　　In a fuzzy control system, as a result of the fuzzy control rules by fuzzy inference decision out of control var

102、iable is a fuzzy subset, it cannot directly control the controlled object, so you also need to take reasonable method to convert them into accurate quantity, so that the best play the decision effect of the result of fuz

103、zy reasoning.</p><p>　　Accurate process is a lot of methods, including MIN - MAX gravity method, algebraic product - addition - gravity method, the weighted fuzzy reasoning method, functional reasoning metho