Elektrik makinaları kontrolünde bulanık mantığın uygulanması

Tez (Yüksek Lisans) -- İstanbul Teknik Üniversitesi, Fen Bilimleri Enstitüsü, 1995Thesis (M.Sc.) -- İstanbul Technical University, Institute of Science and Technology, 1995Günümüzde bilimin amacı daha akıllı insan zekasına, daha benzer sistemler geliştirmektir. Bu yüzden kontrol düzenlerinde kullanılan klasik mantık bu amaç için yetersiz kalmaktadır. Çünkü klasik mantığa göre olaylar "1" ve "0" ya da "var" ya da "yok" olarak değerlendirilmektedir. Hava ya sıcaktır ya soğuktur, bir araba ya hızlıdır ya da yavaştır. Oysa insan aklının işleyişi böylesine katı kurallara bağlı değildir, örneğin, sıcak ve soğuk sıfatları; kişiden kişiye değişebilir, veyahutta kişinin kimi zaman sıcak kimi zamanda soğuk kabul ettiği değerler birbirinden farklı olabilir. Dolayısıyla her şey bir " derecelendirme sorunudur". Bu ilkeyle yola çıkan bulanık mantıkçılar, olayları 0 ve 1 gibi iki katı değer yerine 0 ve 1 arasında değişen birçok esnek değerle açıklamışlardır. (0,1) aralığmdaki değer değişimi bir fonksiyonla verilir, buna üyelik fonksiyonu denir. Üyelik fonksiyonları dilsel kurallarla (büyük, küçük, orta, pozitif büyük ve negatif orta gibi) betimlenirler. Bu dilsel kurallar koşullu önermelerle bulanık mantık yasaları haline getirilirler. Böylece insanın akıl yürütmesine daha benzer bir yöntem elde edilmiş olur. Bulanık mantık, bir çok alanda kullanım olanağı bulmuştur. Klasik denetim kuralları karmaşık dinamik modeller gerektirdiği halde, bulanık mantık daha çok deneyime dayalı, fazla ayrıntılı olmayan modellerle çalıştığından denetim hızı ve niteliği yükselir. Ayrıca bulanık mantık denetimi, sistemin öğrenebilmesine de olanak tanır. Bulanık mantık, elektrik makinaları denetimine de birçok olumlu özellik getirmiştir. Elektrik makinalarının kontrolü, çoğu kez karmaşık modellere dayanır ve parametrelerin değişiminden etkilenir (özellikle asenkron motor). Oysa bulanık kontrolörler kendi özelliklerinden dolayı dinamik model gerektirmezler ve parametre değişimlerinden etkilenmezler. Bulanık kontrolün uygulanabilmesi için öncelikle, sistemi gözetleyerek elde edilen bilgilerle, giriş ve çıkışların bir uzman tarafından bulanıklaştırılması gerekir. Bu arada girişler ve çıkışlar arasında bir kural tablosu kurmak gereklidir. Çok parametreli kontrolörde bazılarının bulanıklaştırılması yeterlidir. Çalışmanın son bölümünde bulanık mantıkla kontrol edilen bir fan sisteminin ve genel bir bulanık kontrol düzeneğinin bilgisayar benzetimleri yapılmıştır.Fuzzy Sets Fuzzy sets were introduced by Lotfi A. Zadeh of Berkeley in 1965 to incorporate concepts such as vagueness into computer systems. In conventional set theory an element either belongs to a set or does not, but in normal human decision making this sharp distinction does not normally exist. For example, to define someone as being "intelligent" is meaningless since there is no definite IQ above which a person is described as intelligent, that is, there are "fuzzy" levels where he could be referred to as "reasonably intelligent", etc. Such blurring of the set boundaries are obtained in fuzzy set theory by defining a membership function of fuzzy set which can be between 0 and 1. In traditional set theory the membership function can only take the values 0 and 1. The Concept of Fuzzy Theory Fuzzy theory is a mathematical theory, and what is called fuzziness takes in one aspect of uncertainty. Fuzziness is the ambiguity that can be found in the definition of a concept or the meaning of a word. For example, the uncertainty in expressions like "old person", "high temperature", or "small number" can be called fuzziness. Up to now probability has been the only uncertainty with which mathematics has worked. The uncertainty of probability generally relates to the occurance of phenomena, as symbolized by the concept of randomness. For example, "it will rain tomorrow" "roll the dice and get a three" have the uncertainty of phenomenological occurances. Randomness and fuzziness differ in nature; that is, they are differrent aspects of uncertainty. For example since the uncertainty of "it will rain tomorrow" comes about because of a prediction made before tomorrow arrives, it will be clarified by the passage of time and the arrival of tomorrow. The uncertainty in " roll the dice and get a three" is also a product of guessing before the roll, and if we actually roll dice and test it, the proposition becomes certain. However, the uncertainty found in "old person" or "high temperature" is not clarified by the passage of time or testing. The ambiguity lies in the meaning of the words, and since it is an essential characteristic of the the words, it always follows them around to some extent. VI Probability theory is said to have been developed in the 17th century, so it has a long history. From the very beginning, engineering made use of its ideas, and they are widely used in the natural sciences. On the other hand, fuzzy theory was only developed about 25 years ago and its use is not yet widespread; but fuzziness expresses a much more everyday uncertainty that is a part of the meaning of words, and words are indivisible from man1 s thinking. All people think and transmit their thoughts and information by means of words. If probability weren' t known to the general public through the announcements of the weather bureau, only people who like to gamble and those preparing for entrance examinations would have anything to do with it. However, everyone is involved with fuzziness, and it is a kind of uncertainty that anyone can understand. If this type of uncertainty could be dealt with mathematically and engineering could make use of it, the effects would be immeasurable. It said that difference between computers, which can only process two-valued information, and people is that the latter can deal with ambiguity, but now this outstanding human ability can be expressed by fuzzy theory, handed over to computers and applied to engineering. Aside from conceptual definitions and the meanings of words, some conceive of fuzziness broadly enough to include things like the uncertainty of people's subjective judgements. The well-known subjectivity factor expresses judgements of one-time phenomenological occurances in terms of probability. This wide definition of fuzziness includes probability-type uncertainty as one type of judgmental uncertainty. The general terms of fuzzy theory that makes use of fuzziness are fuzzy set theory, fuzzy logic and fuzzy measure theory. Fuzzy set theory expresses fuzziness in the narrow sense by means of the concept of sets. Fuzzy measure theory is a theory that handles fuzziness in the wider sense. Fuzzy logic is the concept of fuzzy sets incorporated into framework of multivalued logic. There is also what is called "fuzzy mathematics", standard math into which fuzzy sets and the principles of fuzzy measure. Fuzzy Logic in Control Systems During the past several years, fuzzy control has emerged as one the most active and fruitful areas for research in the applications of fuzzy set theory, especially in the realm of industrial processes, which do not lend themselves to control by conventional methods because of a lack quantitative data regarding the input- output relations. Fuzzy control based on fuzzy logic - a logical system which is much closer in spirit the human thinking and natural language than traditional logical systems. The fuzzy logic controller (FLC) based on fuzzy logic provides a means of converting a linguistic control strategy. A survey of the FLC is present; a general methodology for constructing an FLC and assessing its performance is described; and problems that need further research are pointed out. In particular, the exposition includes a discusion of fuzzification and defuzzification strategies, the derivation of the database vıı and fuzzy control rules, the definition of fuzzy implication, and an analysis of fuzzy reasoning mechanisms. Fuzzy logic, which is the logic on which fuzzy control is based, is much closer in spirit to human thinking and natural language than the traditional logical systems. Basically, it provides an effective means of capturing the approximate, inexact nature of the real world. Viewed in this perspective, the essential part of the FIX is a set of linguistic control rules related by the dual concepts of fuzzy implication and compositional rule of inference. In essence, then, the FLC provides an algorithm which can convert the linguistic control strategy based on expert knowledge into an automatic control strategy. Experience shows that the FLC yields results superior to those obtained by conventional control algorithms. In particular, the methodology of the FLC apppears very useful. When the processes are too complex for analysis by conventional quantitative techiques or when the avaliable sources of information are interpreted qalitatively, inexactly, or uncertainty. Thus fuzzy logic control may be viewed as a step toward a reapproachement between conventional precise mathematical control and human-like decision making, as indicated by. A collection of control rules is a fuzzy algorithm an represents a powerful concept which can be used to provide fuzzy models and fuzzy controllers for use in control engineering. In this way, fuzzy set theory can be used to determine collection of implication statements which casually link input and output fuzzy sets and thereby replace the need for the usual rigorous mathematical models. It should be noted that such relations are only possible if knowledge concerning the process under investigation is available a priori, altough methods that incorporate a learning phase to generate the knowledge base are also becoming available. Assuming that the relations can be formulated, a fuzzy control algorithm can be designed and implemented. Such algorithms are essentially a collection of "IF" and "THEN" statements which use the input/output fuzzy sets, and relations defined, to deduce the control signal necessary to achieve the desired response. It is worth noting that although fuzzy controllers work with fuzzy concepts they have to accept non-fuzzy input variables (from system sensors), and pnduce non- fuzzy control outputs (to drive system actuators). FLC contains four principal components: a fuzzification interface, a knowledge base, decision-making logic, defuzzification interface. 1) The fuzzification interface involves the following functions: a) measures the values of input values, b) performs a scale mapping that transfers the range of values of input variables into corresponding universes of discourse, c) performs thr function of fuzzification that converts input data into suitable linguistic values which may be viewed as labels of fuzzy sets. 2) The knowledge base comprises aknowledge of the application domain and the attendant control goals. It consists of a "data base" and "linguistic (fuzzy) control rule base" a) the data provides necessary definitions, which are used to define linguistic control rules and fuzzy data manipulation in an FLC, vııı b) the rule base characterizes the control goals and control policy of the domain experts by means of a set of linguistic control rules. 3) The decisionmaking logic is the kernel of an FLC; it has the capability of simulating human decisionmaking based on fuzzy concepts and of inferring fuzzy control actions employing fuzzy implication and the rules of inference in fuzzy logic. 4) The defuzzification interface performs the following functions: a) a scale mapping, which converts the range of values of output variables into corresponding universes of discourse, b) defuzzification, which yields a nonfuzzy control action from an inferred fuzzy control action The principal design parameters for an FLC are the following: 1) fuzzification strategies and the interpretation of a fuzzification operatör (fiızzifier) 2) data base: a) discretization/normalization of universes of discourse, b) fuzzy partition of the input and output spaces, c) completeness, d) choice of membership function of a primary fuzzy set; 3) rulebase: a) choice of process state (input) variables and control (output) variables of fuzzy control rules, b) source and derivation of fuzzy control rules, c) types of fuzzy control rules, d) consistency, interactivity, completeness of fuzzy control rules; 4) decision making logic: a) definition a fuzzy implication, b) interpretation the sentence connective and, c) interpretation the sentence connective also, d) definitions of compositional operator, e) inference mechanism; 5) defuzzification strategies and the interpretation of a defuzzification operator (defuzzifier). Fuzzy Logic Applications In Control of Electrical Machines As in many areas, fuzzy logic has found an important place in control of electrical machines. Conventional controllers for electric motors have been developed by means of mathematical models. But in many cases, they can not cope with varying enviroments as a result of load disturbances, system nonlinearity, and changes of parameters etc.. In the worst case, the system may be unstable. For this, it is very important to design a robust controller. Electrical machines (particularly induction machines) have nonlinear characteristics. Therefore, classical control theory can not provide the performance satisfaction. However, fuzzy control developes the control performance, since it is not necessary a dynamic model. An expert must observe the system behaviour and produce a fuzzy control algorithm. IX Some significant reasons can be given to use the fuzzy-set theory and its application. The induction motor is a bilinear many-variable system with some parameters which are nonlinear and depended on the working point. Another parameters are changed slowly by the temperature. The inverter for supplying the induction motor has also a nonlinear characteristic because the switching modes lead to stochastical behaviour and different dead times. The direct control methods for torque and flux-impression are accompanied with influences, each other of the both essential controllers. If two on-off controllers are used the control actions are sharp limited by the given tolerance. The signal processing should be adapted on the characteristics of a quasi-analogue controller and the requirements of a digital equipment for calculation the model of the plant. The fuzzy system realizes the regulation of the motor torque and the stator- flux amplitude preselecting the most advantageous voltage phasor and its pulse-times. The fuzzification, various logical operations and the defüzzification, they are the task for controlling. The results are given by the two voltage phasors and its degrees of membership. The pulse-times of the preselected voltages are produced and the signals for the inverter are generated after defuzification. In another applications; an efficiency optimization method uses a fuzzy controller based on drive measured input power, thus yielding true optimum efficiency operation with fast convergence. At steady state light load condition, a fuzzy controller adaptively decrements the excititation current on the basis of the measured input power such that, for a given torque and speed, the drive settles down to the minimum input power; i.e., operates at maximum efficiency. Fuzzy logic has reached an important role in control of other kinds of electrical machines, such as, switched reluctance motor, step rotor, brushless d.c. motor etc. However, structure of a fuzzy controller is universal an compatible for this reason. It can be applied to any drive system. This study has five main parts. In the first part, it has been given an introduction to fuzzy logic. In the second and third part, mathematical relations of fuzzy logic and fuzzy logic controller have been explained. In the fourth part, fuzzy control techniques for electrical machine drives have been evaluated by using several examples. At the final part of this study. Two computer programs have been developed.Yüksek LisansM.Sc