Konoval VA Control systems with fuzzy logic

RUS | UKR | | DonNTU > portal Masters DonNTU

Magistr DonNTU Konoval Vladimir Anatolevich

Vladimir Konoval

Faculty of Computer Science and Informatics
Specialty: Computer systems and networks

Subject graduation work:

Control systems with fuzzy logic

Leader of work: Sergei Kovalev

Summary on exhaust work

Introduction

Today, the term fuzzy set and fuzzy logic are not so well known, as, for example, in the mid-eighties. Then the phrase "fuzzy logic" is not literally get away from the pages of various publications - from the little-known highly specialized to the mass of scientific and well-known that the term "research" at the mention of their names had to express something shameful. So was the fact that the words of greedy Americans call hype. Again, it is clear - to understand the hype is possible only after its completion, as happened with all that had anything to be related to a fuzzy term of fuzzy. By the mid-90 s promising technology started to criticize. First, an analysis of citations in scientific papers has revealed a catastrophic decline in interest in academics. Then answered the practice - like the word fuzzy marked the most "blind-alley" section in the programming. And, by the beginning of a new tisyachilittya, many of them started a prediction that the fuzzy logic - it is something once very fashionable and now completely forgotten. It certainly happens, it is actually true, and at the same time not so. When superimposed on the "fuzzy logic" as the theory and technology solutions to all problems have not been confirmed expectations - just as, if not previously verified and not confirmed in the future all the same expectations of the most advanced treatments for all diseases. But the theory and practice of that in any case is not affected. Rather the contrary. On the one hand, fuzzy logic is found so well-defined area of use, made possible the emergence of powerful modeling tool, allowing you to hide a lot of nontrivial complex mathematical operations for the user-friendly interface and severe problem graphics-oriented metaphors. On the other hand, basic mathematical operations of fuzzy logic are so clearly defined that they have long and successfully implemented "in iron" (more precisely, in the systems commands), mass-produced single-chip microcontrollers.

Relevance

Now becoming increasingly important use of expert systems for solving three-dimensional, it is difficult formalized tasks in various subject areas. These tasks are characterized, as a rule, or the lack of formal complexity algorithms for the solution, the incompleteness and vagueness of the initial information, unclear goals are achieved. These features make it necessary to use in addressing these problems of knowledge derived from the human expert in a subject area, and the development of expert systems for the collection and management of knowledge, which shall decide on the best way to achieve the conditions of incompleteness and fuzziness.
Human expert knowledge about problem solving in conditions of incompleteness, ambiguity of background information and goals are achieved, are also unclear. For their formalization is now successfully applied the apparatus of the theory of fuzzy sets and fuzzy logic. Fuzzy concepts in this case are formalized in the form of fuzzy and linguistic variables, and the vagueness of action in decision-making - in the form of fuzzy algorithms. Expert systems capable formalizovivat obscure information and process it as part of fuzzy algorithms are called fuzzy expert systems.
Great relevance at the present time is the use of fuzzy expert systems for solving problems of modeling in the field of geology and development of oil fields. The main characteristic of the problems solved in this subject area is the uncertainty, ambiguity and incompleteness of knowledge about the field. Knowledge used by geological experts to solve problems in those subject areas, often intuitive and subjective, largely due to the fact that geologists are far from fully understood set of processes occurring during the development of oil fields. On the other hand, many geological deposits rates in principle can not be determined precisely because of its nature. For example, the boundary of the deposit - a perfect object, it can not be defined clearly by its nature, she blurred, vague and fuzzy.
Set the main goals reached during the development of oil fields - extracting resources from the deposits of maximum profit, with a high oil recovery, with minimal resource and financial costs, with minimal loss of environmental - are vague, unclear, vague. They can not be achieved fully and changing over time.

Number of staff: 4 Cycle repetition 0.5mls The blue line outlines the result set (fuzzy AND between multiple 5 and 8)

Number of staff: 4 Cycle repetition 0.5mls The blue line outlines the result set (fuzzy OR between multiple 5 and 8)

Scientific novelty

In defazifikatsiey defined as a procedure transforming the fuzzy values derived from fuzzy inference, in the clear. This procedure is necessary in those cases that require interpretation of the fuzzy conclusions specific distinct values, that is, when based on the membership function is necessary to determine for each point in Z numbers.
There is currently no systematic procedure for choosing a strategy defazifikatsii. In practice, often use the two most common methods: the method of center of gravity (CG - tsentroidny), the method of maximum (MM).
Of the two most frequently used strategies defazifikatsy, MM strategy gives better results for the transitional regime, and CT - in the steady state for the lesser mean square error.
The idea is to modify the methods defazifikatsii, which leads to optimal block zbilshennya defazifikatsii on structural scheme and this in turn leads to an increase in performance throughout the facility. And in our time is very popular saying - time - money ... and besides big bitter. And if we virobnitsvo cheaper ... so here are great prospects and a weight of scientific value ...

Conclusion

The work has been studied a lot of material on the foundations of control systems based on fuzzy logic.
basic idea, which is used in the SNL, is to enter "the experience of the expert" (the human operator the person receiving (LPS)) in the development of the scheme, managing some of the dynamic process. Nowadays SNL distributed very widely, and find their purpose in the various fields of science and technology. This is due to the fact that the control system with fuzzy logic provide very significant gain in time on the existing microcontrollers. Here's an example:
There is one family of microcontrollers that have become today a classic mass-produced affordable cars fuzzy logic. Naturally, we are talking about the famous family of HC12 company Freescale (a former semiconductor division of Motorola). The system commands HC12 implemented such a unique design, such as, MEM, and WAV, in fact, is odnokomandnoyu implementation procedures fazifikatsii and defazifikatsii. In addition, HC12 supports two teams engaged in intermediate stages of the mechanism of fuzzy inference. Add to that the four highly specialized machines nearly a dozen teams "fuzzy-oriented" instruction need not be surprised that the HC12 so loved and respected manufacturers of systems, embedded in various fields. After all, the overseers of this family of fuzzy problems overtake the more expensive and versatile 32 bit chips are not a few, but in tens and hundreds of times. But all this is possible only when the correct use and development of elements of fuzzy logic.

Literature

1. Peter Bauer, Stephan Nouak, Roman Winkler "Fuzzy Logic" December 4, 1996
2. Kruglov VV "Fuzzy logic and artificial neural networks", 2002 - 382c
3. Featured articles Lotfi Zadeh
Kosko B. Fuzzy systems as universal approximators // IEEE Transactions on Computers, vol. 43, No. 11, November 1994. - P. 1329-1333.
Леоленков А.В. Нечеткое моделирование в среде MATLAB и fuzzyTECH. - СПб., 2003.
Рутковская Д., Пилиньский М., Рутковский Л. Нейронные сети, генетические алгоритмы и нечеткие системы. - М., 2004.
Cordon O., Herrera F., A General study on genetic fuzzy systems // Genetic Algorithms in engineering and computer science, 1995. - P. 33-57.