Social and Engineering Applications

Part I. Philosophical and Mathematical Fundamentals of Fuzzy System Design

Part II. Fuzzy System Design for Social Applications

Part III. Fuzzy System Design for Engineering Applications

This book goal is constructing bridges and filling gaps. However, one should not think that it is about civil engineering. The gaps considered are not real although they do exist.

The first gap which we would like to overcome is one between a theory and practice, between academia on one side and a community of practitioners on another. Unfortunately, this gap becomes more and more obvious and now it shows up even at the consideration level. That is why we concentrate our discussion on design problems with a design being a top of a research and development process, its last point and a gate to an implementation, where all methods applied and decisions made at the preceding stages are actually tested. On the other hand, a lion's share of this book is occupied by the sections devoted to the design of fuzzy systems for solving practical problems.

This book considers applications in both social and engineering fields and this is another bridge which it is meant to build between these two application areas and corresponding communities, which can use it to learn from each other. Engineers, researchers and students working in fuzzy engineering may learn about social impact of fuzzy system applications and acquire some knowledge about an area, which undoubtfully has an unlimited potential for a future development. On the other hand, social scientists and students may be given a hint how to apply real working methods and tools of fuzzy engineering in solving their problems.

Fuzzy sets and systems theory was introduced by Prof. Lotfi Zadeh in 1965. Not at the least degree attempting to decrease his merit in an invention, development and promotion of a new approach, we would like to mention that this overture was prepared by a previous history of science and, first of all, logic science. And one of the forces initiating the birth of fuzzy logic was some famous ancient paradoxes - the problems which could not be solved by classical logic. The history of fuzzy logic is filled with some paradoxes as well. One of the main paradoxes is that despite the founders of this theory initially expected main applications to appear in social sciences and large organisational systems design, most of real applications until now have been developed in an engineering field. Nowadays the problem how to share the experience and combine the developed methodologies of an engineering design with the great potential of social applications has emerged.

The volume consists of three parts. The first one includes chapters on philosophical and mathematical basics of fuzzy system design. Being a solid theoretical fundamental it serves as a bridge between two other parts considering social and engineering applications. One should note, however, that some authors are very successful in constructing bridges in their papers which include some theoretical developments and application problem solutions together.

The book starts with a discussion of the philosophical foundations of fuzzy logic given in the chapter "Fuzzy Leadership: Dancing with Organisational Reality" by P. Duignan. The paper proposes to apply the "fuzzy logic frame" as an attempt to gain deeper understanding of social reality, especially the nature of leadership. Different aspects of natural and social systems and applications of fuzzy systems for leadership and management are discussed based on the concept of a "cosmic dance of energy". In order to understand why and how the individuals join together and share in a group or organisation the metaphor of "free-form dance" is examined. The problems of the language description and interpretation are investigated. This paper should be interesting for any reader with either "social" or "engineering" background.

The chapter "Multi-Criteria Optimisation - an Important Foundation of Fuzzy System Design" by H.T. Nguyen and V. Kreinovich considers a problem which is quite common in different research areas and typical in design. How to design a control device with both good steady-state response and excellent transient characteristics? How to increase budget revenues and decrease taxes? The traditional way of trying to achieve several goals simultaneously is their combination into one. Besides the obvious fact that one achieves neither objective, it is not often clear which combination to apply. As authors claim, fuzzy logic proposes a natural tool how to handle multicriteria optimisation. The paper proposes the mathematical fundament for a number of fuzzy logic based optimisation methods developed until now and proves that fuzzy logic application is really "natural" way of design.

Another way of decision making in a fuzzy environment is proposed by V. Korotkih in the chapter "A Mathematical Framework for Human Decision Making as an Integrated part of the Whole". This paper considers a model of human collective decision making. The web of integer relations and structural complexity is proposed as a mathematical framework to model a problem situation in which each individual decision is considered in its relation to others or as a part of the whole process. The process of decision making in this case is supposed to be a multi-step and adaptive. This web is used later to study properties of strategies of the optimal decision search and formulation of the optimal rules. As one may see this model can be definitely considered as a mathematical base for the design of both social and engineering systems.

The study of decision making continues in the next chapter "Fuzzy Decision Making in Design on the Basis of the Habituality Situation Application" by A.E. Gorodetsky. Although this paper considers a mathematical formalism, it has a practical goal: to improve an efficiency of expert systems. The way how to reach this goal is quite obvious - one has to teach an expert system how to apply some tricks used by a human expert. One of those tricks is the decision making based on the so-called habituality situation application. If an environment situation at the decision moment is similar to the situation observed previously for which a good decision is known, there is no reason to waste time and other resources by looking for another decision. The chapter describes how to determine the habituality situation and to decide if the situation is similar enough to apply the well-known decision.

The same idea of creating intelligent artificial tools by including some features previously specific to the human designers is exploited in the next chapter "Fuzzy Logic Applications in CAD" by B. Pham. As one can see from the title it is associated with the computer aided design. However, the author's main idea is rather non-standard for contemporary CAD systems. In order to improve the CAD system quality the paper proposes to include into CAD systems some methods featuring a human-like design way. These methods are based on the application of aesthetic factors and their fuzzy models. It will allow to expand CAD applications from traditional limited engineering and architectural areas to other types of design such as sculpting or industrial design, and probably further to social system design, where aesthetic factors play a more important role. And definitely these CAD systems will be introduced with some characteristics, not typical for computers, like creativity and artistic abilities

In this book a number of articles of the Part II illustrate various social applications of fuzzy logic. An original framework for fuzzy system design related to modern participatory management based on fuzzy logic is developed; it is shown also how fuzzy system design helps to unlock some destructive cycles continuously formed in complexity of human and organisational ecology.

R. Woog, V.Dimitrov and L.Kuhn-White in their paper "Fuzzy Logic as an Evocative Framework for Studying Social Systems" share their experience of fuzzy logic applications in social system research. This chapter answers a question why fuzzy logic is useful in social research. The authors emphasise a crucial role for fuzzy logic in social inquiry and such its fields as a conversation mapping, a heuristic pattern formation, an emergence of meaning, a multi-layered interpretation, a study of temporality and non-foundations thinking. One of the major contributing strengths of a fuzzy logic approach to social inquiry is its ability to cope with multiple constructions of reality, as well as other forms of circumstantial and constructed complexity. In the application to research/inquiry methodology, fuzzy logic helps to avoid absolute statements and to deny the legitimacy of all dichotomies, to embrace ambiguity and ambivalence, to seek freedom from the myopia of hyper-determined research projects and formulations, and to be aware of the tentativeness of causal explanations.

The discussion of management problems is continued in the paper "Fuzzy Logic and the Management of Social Complexity" by V.Dimitrov and K.Kopra. The presentation is concentrated on an application of fuzzy logic to management of paradoxes inherent in social systems and their effects as well as the chaos of social life. The role of fuzzy logic in understanding and healing some sociol-political maladies of contemporary societies is paid a special attention. Among the maladies considered are: a false face of representative politics, a misconception of liberty and equality as mutually exclusive, a destructive competition based on short-sighted economic rationalism, a disconnection of economy from society, a no-alternatives strategy, a false concept of a personal identity. One can see how important these problems are for a modern society. Regardless one's political and philosophical points of view, the chapter provides a reader with a research tool for a social system understanding and analysis.

The next chapter "Using Fuzzy Sets to Extend Holland's Theory of Occupational Interests" by M.Smithson and B.Hesketh considers a fuzzy system application to enhance a famous in psychology typological theory of careers and career choice introduced by Holland. This application has several advantages, such as the possible use of different types of analysis which both sharpen and extend the meaning of consistency and congruency, the focus on the relation of the individual to different groups, and better handling the information contributed by measuring avoidance reactions. A large part of the paper is devoted to the report of the empirical studies in which the fuzzy sets and systems methodology was applied and to the analysis of the results achieved.

Another psychological application is considered in the next chapter "A Model for Fuzzy Personal Construct Psychology" by A. Anderson. The author attempts to fuzzify another well-known psychological theory - the personal construct psychology by G. Kelly. Personal constructs are a means of discrimination between items (usually called elements) in terms of similarity and contrast. Clearly, this research subject could be important for specialists in different areas. The paper presents a mathematical model which combines the essence of the personal construct psychology with fuzzy logic. The proposed methodology is incorporated into a software package FUZZYGRID and a case study is considered.

Next chapter "On the Topology of Uncertainty" by D.McNeil and V.Dimitrov investigates the role of uncertainty in a real life and the fuzzy logic application to model it. The social framework of fuzzy logic presents a perspective, that makes it possible to understand, how our current world order works. Despite ecological disasters, violence, injustice, poverty and wars, there is some kind of order that keeps humanity resisting the forces of distraction. Paradoxically, the driving mechanism of this order is uncertainty, embedded in our individual and social life. It is the endeavour for dealing with this uncertainty and trying to grasp and manage its multifarious aspects that keep people looking for consensus and togetherness, although it seems difficult to be sometimes. Fuzzy logic is a way of thinking that is responsive to human zeal to unveil uncertainty and deal with social paradoxes emerging from it.

Different fuzzy logic applications in engineering have become well known. Various software and hardware design tools have been developed and applied successfully. So an engineering part of the book contains a number of papers, devoted to the description of fuzzy engineering design methodologies. In order to share the experience gained we have selected papers describing not the application result only but the way how this result has been obtained, that is explaining the design procedures. The papers chosen consider mainly not the theory of fuzzy system design but the practical aspects of design.

The chapter, which opens up a collection of the papers devoted to the problems of fuzzy control system design and applications, "Fuzzy Controller Design for Different Applications: Evolution, Methods, and Practical Recommendations" by L.Reznik briefly describes a short history of fuzzy controller design underlining an interaction of different approaches in fuzzy system design. This competition between different methods has become a driving force for further improvement and development of new synergetic design methodologies. Not attempting to propose a comprehensive mathematical theory of design, the paper aims at providing some advice which way of design to apply, and how to choose the structure and parameters of a fuzzy controller. These recommendations could be useful for designers of different areas fuzzy systems.

L.Koczy and D.Tikk in their paper "Approximation of Transfer Functions by Various Fuzzy Controllers" are trying to expand a very popular in control engineering as well as simple mechanism of transfer functions to a domain of any fuzzy system. Or at least, the most widely used fuzzy systems with a reasonably low computational complexity can be included. If proved, this method will allow to construct a bridge to an application of very well developed analytical and design methodologies to the artificial intelligence systems.

Next chapter "Fuzzy Performance Indicators for Manufacturing Processes" by L.Berrah, G.Mauris, L.Foulloy, and A.Haurat talks about a control of manufacturing processes. In addition to the accounting measures and the financial reports, supervisors usually require performance indicators for the "physical" evaluation of the manufacturing process evolution, leading so to a more reactive control. According to pre-set objectives, the indicators evaluate the results achieved, and measure the extent of any detected drift or malfunctioning . These indicators can be formed on the base of both technical and human generated data. When the measures are effected by physical sensors, they are numerical and eventually pervaded by errors, often described by a Gaussian distribution, while the measures related to subjective estimates are often acquired by human operators, who describe them as linguistic variables. Here an engineering problem becomes an organisational problem. Can fuzzy technology help to solve both of them together?

The chapter "Advanced Neuro-Fuzzy Engineering for Building Intelligent Adaptive Information Systems" by N.K Kasabov is a brilliant example of an application of engineering methodologies to a new domain of information systems. The paper introduces a new architecture of fuzzy neural networks (FNN). These FNN have been suggested and applied by several authors for learning and tuning fuzzy rules and solving classification, prediction and control problems while applied mainly in an engineering area. In this paper the author proposes a new architecture of FNN, called FuNN and develops a general methodology for building modular FuNN-based adaptive intelligent systems. Interesting enough, the possible applications considered in the chapter include adaptive intelligent speech interfaces to databases and an adaptive prediction of chaotic time-series with financial applications. The results achieved demonstrate how strong the "engineering factor" has become in business and information management applications.

The authors of the next chapter "Linguistic Variables in Surveys of Intentions to Purchase Energy Efficient Equipment", W.Mielczarski, G.Michalik, M.E. Khan are engineers. However, the topic of their research and their results can cause a strong interest of "social" scientists, businessmen, market researchers, and even politicians. They are applying linguistic variables and scales to analyse a customer's purchase intentions. By the way, the same method seems to be suitable for forecasting an electorate's voting and for any type of forecasting based on customer surveys. The approach proposed can be a tool for the analysis of linguistic terms, assigned subjective probabilities and sensitivity studies.

Although the application considered in the next chapter "A Universal Approach to Adaptive Fuzzy Logic Controller Design With an Application to a Power Generator Excitation Control" by O.Ghanayem and L. Reznik is a traditional power engineering area of a generator control, the proposed design methodology has a universal nature. The paper introduces a new method for on-line tuning of fuzzy controller scaling factors. An importance of the scaling factors right choice in design is very difficult to underevaluate due to their influence on fuzzy control system stability and performance. This is why the problem has been investigated for some time and some results have been achieved. Main features of this paper are the adaptation of both input and output factors simultaneously and on-line tuning. These features allow to apply this method in design of different fuzzy systems including control systems for the plant with unknown models.

The next chapter "Fuzzy Logic Applications in Diagnosing Mechatronic Systems" by T.Rauma and M.Kurki is exploiting a tolerance of fuzzy logic models to uncertainty. The proposed methodology includes constructing of fuzzy models, modifying them, and applying them in solving different problems. It allows to apply fuzzy models in solving complex problems of a fault detection and diagnosis and building an interface with an operator. A few examples of real life applications in diagnosis of mechatronic systems are presented.

The last chapter "The Application of Fuzzy Logic Techniques for the Enhancement of Handoff Characteristics in Digital Cellular Mobile Communication Networks" by P.S.K. Leung considers fuzzy system design application in the area which has been marked by phenomenal growth and rapid technological advances during last two decades. In cellular engineering the problem addressed in the mathematical part of the book is quite typical. There are two conflicting criteria to be minimised: a number of unnecessary handoffs between basestations and a number of lost calls. As a solution the fuzzy system is designed to improve the handoff characteristics of a future generation of micro- and pico-cellular wireless communication systems. The paper investigates such design problems as a choice of inputs, rules, membership functions and inference methods.

As one can see different researchers from various countries and continents have made their generous contribution into this volume. The editors would like to express the great appreciation of their work and research results and thank all those who have made this publication possible.

L.Reznik, V.Dimitrov, J.Kacprzyk

Melbourne, Sydney, Warsaw June 1997

site description

Name of link

Copyright © 1997 Leon Reznik
This Home Page was created by WebEdit, Friday, 29 August 1997
Most recent revision Friday, 31 August 1997