Showing papers on "Neuro-fuzzy published in 1991"

neural networks and fuzzy systems a dynamical systems approach to machine intelligence

Abstract: This work combines neural networks and fuzzy systems, presenting neural networks as trainable dynamical systems and developing mechanisms and principles of adaption, self-organization, convergence and global stability. It includes the new geometric theory of fuzzy sets, systems and associated memories, and shows how to apply fuzzy set theory to adaptive control and how to generate structured fuzzy systems with unsupervised neural techniques.

Neural-network-based fuzzy logic control and decision system

Abstract: A general neural-network (connectionist) model for fuzzy logic control and decision systems is proposed. This connectionist model, in the form of feedforward multilayer net, combines the idea of fuzzy logic controller and neural-network structure and learning abilities into an integrated neural-network-based fuzzy logic control and decision system. A fuzzy logic control decision network is constructed automatically by learning the training examples itself. By combining both unsupervised (self-organized) and supervised learning schemes, the learning speed converges much faster than the original backpropagation learning algorithm. The connectionist structure avoids the rule-matching time of the inference engine in the traditional fuzzy logic system. Two examples are presented to illustrate the performance and applicability of the proposed model. >

Fuzzy systems theory and its applications

Abstract: Read more and get great! That's what the book enPDFd fuzzy systems theory and its applications will give for every reader to read this book. This is an on-line book provided in this website. Even this book becomes a choice of someone to read, many in the world also loves it so much. As what we talk, when you read more every page of this fuzzy systems theory and its applications, what you will obtain is something great.

Fuzzy Expert Systems

Abstract: PREFACE. FUZZY EXPERT SYSTEMS THEORY. The Evolution From Expert Systems to Fuzzy Expert Systems. General Purpose Fuzzy Expert Systems. Inferences With Inaccuracy and Uncertainty in Expert Systems. On the Representation of Relational Production Rules in Expert Systems. Reduction Procedures for Rule-Based Expert Systems as a Tool for Studies of Properties of Expert's Knowledge. The Physiology of the Expert System. On the Processing of Imperfect Information Using Structured Frameworks. Fuzzy Linguistic Inference Network Generator. Advances in Automated Reasoning Using Possibilistic Logic. Fuzzy Associative Memory Systems. APPLICATIONS OF FUZZY EXPERT SYSTEMS. The Role of Approximate Reasoning in a Medical Expert System. Fess: A Reusable Fuzzy Expert System. Design for Designing: Fuzzy Relational Environmental Design Assistant (FREDA). On the Design of a Fuzzy Intelligent Differential Equation Solver. MILORD: A Fuzzy Expert Systems Shell. Medical Decision Making Using Classification Techniques for Establishment of Knowledge Bases. Fuzzy Expert Systems for an Intelligent Computer-Based Tutor. Expert System on a Chip: An Engine for Approximate Reasoning. A Probabilistic Logic for Expert Systems. COMEX-An Autonomous Fuzzy Expert System for Tactical Communications Networks.

Fuzzy modeling using generalized neural networks and Kalman filter algorithm

Abstract: We propose a new approach to build a, fuzzy inference system of which the parameters can be updated to achieve a desired input-output mapping. The structure of the proposed fuzzy inference system is called generalized neural networks, and its learning procedure (rules to update parameters) is basically composed of a gradient descent algorithm and Kalman filter algorithm. Specifically, we first introduce the concept of generalized neural networks (GNN's) and develop a gradient-descent-based supervised learning procedure to update the GNN's parameters. Secondly, we observe that if the overall output of a GNN is a linear combination of some of its parameters, then these parameters can be identified by one-time application of Kalman filter algorithm to minimize the squared error, According to the simulation results, it is concluded that the proposedl new fuzzy inference system can not only incorporate prior knowledge about the original system but also fine-tune the membership functions of the fuzzy rules as the training data set varies.

Hierarchical fuzzy control

Abstract: In a conventional rule based fuzzy control system, the rules are of the following form: if (condition) then (action), and all rules are essentially in a random order. The number of rules increases exponentially as the number of the system variables, on which the fuzzy rules are based, is increased. In this paper, the rules are structured in a hierarchical way so that the total number of rules will be a linear function of the system variables. The hierarchical fuzzy control algorithm developed in this paper is applied to control the feedwater flow to a steam generator of a power plant. The simulation results show that the hierarchical fuzzy controller yields superior performance over the conventional PID controller.

Fuzzy min-max neural networks

Abstract: A supervised neural network classifier using a combination of min-max hyperboxes and fuzzy logic is described A min-max hyperbox and its membership function define a fuzzy set Each class in the neural network is a collection of labeled hyperboxes (fuzzy sets) The degree to which an input pattern belongs to a class is determined by the membership function of the winning hyperbox Using multiple hyperbox fuzzy sets to form classes allows arbitrary numbers and shapes of classes and their respective class boundaries The min-max classification learning procedure requires only a single pass through the data and allows online learning The author describes how the fuzzy min-max classifier is implemented as a neural network, explains how min-max classes are produced, and provides two examples of operation >

Interpretation of Artificial Neural Networks: Mapping Knowledge-Based Neural Networks into Rules

Abstract: We propose and empirically evaluate a method for the extraction of expert-comprehensible rules from trained neural networks. Our method operates in the context of a three-step process for learning that uses rule-based domain knowledge in combination with neural networks. Empirical tests using real-worlds problems from molecular biology show that the rules our method extracts from trained neural networks: closely reproduce the accuracy of the network from which they came, are superior to the rules derived by a learning system that directly refines symbolic rules, and are expert-comprehensible.

On fuzzy neuron models

Abstract: A contribution to the theoretical development of fuzzy neural network theory is presented. Three types of fuzzy neuron models are proposed. Neuron I is described by logical equations of 'if-then' rules; its inputs are either fuzzy sets or crisp values. Neuron II, with numerical inputs, and neuron III, with fuzzy inputs, are considered to be simple extensions of non-fuzzy neurons. A few methods of how these neurons change themselves during learning to improve their performance are also given. The application of the non-fuzzy neural network approach to fuzzy information processing is briefly discussed. >

Fuzzy associative memories

Abstract: Mappings between fuzzy cubes are discussed. This level of abstraction provides a surprising and fruitful alternative to the propositional and predicate-calculas reasoning techniques used in expert systems. It allows one to reason with sets instead of propositions. Discussed here are fuzzy and neural function estimators, neural vs. fuzzy representation of structured knowledge, fuzzy vector-matrix multiplication, and fuzzy associative memory (FAM) system architecture.

Fuzzy clustering algorithms based on the maximum likelihood principle

Abstract: A number of hard clustering algorithms have been shown to be derivable from the maximum likelihood principle. The only corresponding fuzzy algorithm are the well known fuzzy k-means or fuzzy isodata of Dunn and its generalizations by Bezdek and by Gustafson and Kessel. The authors show how to generate two other fuzzy algorithms which are analogous of known hard algorithms: the minimization of the fuzzy determinant and of the product of fuzzy determinants. By comparison between the hard and fuzzy methods it appears that the latter yield more often the global optimum, rather than merely a local optimum. This result and the comparison between the different algorithms, together with their specific domains of application, are illustrated by a few numerical examples.

Design of a cart-pole balancing fuzzy logic controller using a genetic algorithm

Abstract: Scientists at the U.S. Bureau of Mines are currently investigating ways to combine the control capabilities of fuzzy logic with the learning capabilities of genetic algorithms. Fuzzy logic affords a mechanism for incorporating the uncertainty inherent in most control problems into conventional expert systems. Although fuzzy logic-based expert systems have been used successfully for controlling a number of physical systems, the selection of acceptable fuzzy membership functions has generally been a subjective and time consuming decision. In this paper, high-performance fuzzy membership functions for a fuzzy logic controller that manipulates a mathematical model simulating a cart-pole balancing system are selected using a genetic algorithm, a search technique based on the mechanics of natural genetics. The membership functions chosen by the genetic algorithm provide for a more efficient fuzzy logic controller than membership functions selected by the author for the cart-pole balancing problem. Thus, genetic algorithms represent a potentially effective and structured approach for designing fuzzy logic controllers.© (1991) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

Fuzzy rule extraction from a multilayered neural network

Abstract: A fuzzy rule extraction method from a multilayered neural network is proposed to realize the multiplicative merits of neural and fuzzy theories. The proposed method uses a fuzzy-neuron network structure which includes input and output layers that convert input signals to membership values. An evaluation index, called the 'casual index' can evaluate the weights which are acquired by learning and can translate the internal knowledge of the network into fuzzy rules. Simulations using test data whose relationships are known a priori demonstrated the ability of the proposed method to extract fuzzy rules properly from a multilayered neural network. >

Fuzzy logic control compared with other automatic control approaches

Abstract: The authors present a comparison of fuzzy control algorithms with those of conventional control approaches (PID, optimal) in the case of a linear dynamic process control. They first give an overview of the different controllers used and examine the sensitivity of the fuzzy logic controller to design parameters. A comparative study between three types of controllers with regard to performance and robustness is then presented. It is concluded that the main interest of the fuzzy controller lies in its easy implementation. The expertise and experience of the human operator make it possible to elaborate a set of rules to control the system without requiring the computation of any mathematical model. For simple systems, a real-time adaptation is eased by using quantizations and look-up tables, which result in very short development and perfecting times. A fuzzy controlled system shows good results in terms of response time and precision. Moreover, such a controller allows a good robustness mainly with respect to modifications of process parameters as well as various disturbances on the control. >

A framework for analysis and synthesis of fuzzy linguistic control systems

Abstract: Fuzzy Linguistic Control(FLC) may be viewed as a knowledge based control strategy that can be used when the dynamic characteristics of the plant or the associated control objectives are not sufficiently well posed to warrant the, at least immediate, application of conventional control techniques. Alternatively, in these situations one may opt for a heuristically designed control strategy that relies on empirically acquired knowledge regarding the operation of the process. This knowledge cast into linguistic or rule based form constitutes the core of a fuzzy linguistic control system. In other words, the control law, instead of being stated as an analytic function of the process output (or state for that matter), is composed of condition $\to$ action rules that capture, in an approximate sense, the empirical knowledge or know-how necessary to effectively control the process. This heuristic approach to control design raises important issues regarding the stability and reliability of fuzzy linguistic control systems. In this connection, the purpose of this dissertation is twofold. First, we present an approach to the analysis of stability of fuzzy linguistic control systems that is based on Lyapunov's Direct Method and as such offers sufficient conditions, which if met, guarantee global asymptotic stability of the system under consideration. Second, we will present an alternative framework for analysis, as well as synthesis, of fuzzy linguistic control systems that is intended to bridge the gap between this type of control strategy and conventional control techniques. Specifically, by parametrizing the characteristic functions of fuzzy subsets describing the linguistic terms used in the definition of the control rules, we develop an analytic formulation of the control scheme, which is then used to derive sufficient conditions for asymptotic stability of the closed loop dynamic system operating under fuzzy linguistic control. In this process, we will also establish a connection between fuzzy linguistic control and a class of nonlinear control schemes, namely those with piecewise linear characteristics. Furthermore, we present a new interpretation of fuzzy linguistic control that helps explain how certain types of nonlinearities, such as asymmetric response characteristics, can in effect be canceled by appropriate formulation of control rules. Finally, we will suggest how fuzzy linguistic control, in general, can be used to deal with other commonly occurring nonlinear behavior in industrial processes such as actuator saturation, nonlinear sensor behavior, and/or operating regime dependent variations in the response characteristics of the process.

Fuzzy set representation of neural network classification boundaries

Abstract: In neural network classification techniques, the uncertainty of a new observation belonging to a particular class is difficult to express in statistical terms. On the other hand, statistical classification techniques are also poor for supplying uncertainty information for new observations. The use of fuzzy sets is a promising approach to providing imprecise class membership information. The monotonic function neural network is a tool that can be used to develop fuzzy membership functions. This research suggests that a multiarchitecture monotonic function neural network can be used for fuzzy set representation of classification boundaries in monotonic pattern recognition. >

Fuzzy logic computer.

Abstract: FUZZY LOGIC COMPUTER A RISC computer is specially adapted in its hardware and operating procedures for operating systems based upon sets of logical statements utilizing fuzzy logic procedures. The computer may operate either stand-alone or embedded in a host computer, and supplies system control signals derived from measurements of system operating conditions and operating knowledge expressed in fuzzy logic form. A core processor of the computer performs logical evaluations and mathematical manipulations, and performs other control and memory management functions. A key feature of the logical evaluation hardware approach is the evaluation of logical conclusions by their areas and moments determined prior to program execution, permitting much faster computation of command signals than previously possible. Communication to the controlled system or a host computer is through an arbiter that manages the data flow in conjunction with a data memory.

An improved design procedure for fuzzy control systems

Abstract: The purpose of this paper is to introduce simpler and more effective method in designing fuzzy controllers. A scale for partitioning the universe of discourse and choosing the appropriate fuzzy set shapes for the control variables is introduced and is termed “fuzzimetric arcs”. Knowledge engineering techniques are used to obtain process information which is interpreted using systematic analysis as an aid to the design of fuzzy logic controller. This has been illustrated by a manufacturing system example in the form of a welding application which provides a guide to the reader who is unfamiliar with such techniques.

A Theory of Mass Assignments for Artificial Intelligence

Abstract: We present a theory of mass assignments for evidential reasoning under uncertainty which allows for fuzzy[15, 16, 17], probabilistic and incomplete probabilistic specifications[1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11]. The theory is applicable to fuzzy control, expert systems, decision support systems, knowledge engineering and represents a general theory of uncertainty in AI.

Basic Issues on Fuzzy Rules and Their Application to Fuzzy Control

Abstract: Fuzzy logic controllers have encountered an extraordinary success in a great variety of industrial applications in the last few years, especially in Japan. The principle of fuzzy controllers, first outlined by Zadeh[31] and then successfully experimented by Mamdani and Assilian[20], consists of synthesizing a control law for a system from fuzzy rules, usually provided by experts, which state the action(s) to do in typical situations, in contrast with the standard approach to automatic control which requires a model of the system to control. Each rule more or less applies to a fuzzy class of situations and an interpolation operation is performed between the conclusion parts of the selected rules, on the basis of the degrees of compatibility between the condition parts of these rules and the current situation encountered by the system. The reader is referred to Mamdani[19], Sugeno[24] for introductions and to Lee[16] and Berenji[5] for surveys. The basic methodology of fuzzy logic controllers was empirically developed in the late seventies and early eighties and has not changed much since. Recently, a revival of rather theoretically-oriented studies has been observed in order to build a strong methodology for fuzzy logic controllers. Thus the analytical comparison between a fuzzy controller and a proportional-integral controller[30], the limit behavior of fuzzy controllers[6], the stability of fuzzy controllers[27], [26], adaptive techniques for fuzzy controllers, e.g. [22]; [1]; [13], and the use of neural network methods for learning fuzzy rules and implementation issues[17] [28] have been discussed.

Interactive Fuzzy Optimization

Abstract: 1 Introductory Sections- Fuzzy set theory and modelling of natural language semantics- A survey of fuzzy optimization and mathematical programming- 2 Fuzzy Optimization: General Issues and Related Topics- Minimizing a fuzzy function- A concept of optimality for fuzzified mathematical programming problems- Some properties of possibilistic linear equality systems with weakly noninteractive fuzzy numbers- Fuzzy preferences in linear programming- Implication relations, equivalence relations and hierarchical structure of attributes in multiple criteria decision making- Uncertain multiobjective programming as a game against nature- Approaching fuzzy integer linear programming problems- Interactive bicriteria integer programming: a performance analysis- Interactive approaches for solving some decision making problems in the Czechoslovak power industry- 3 Issues Related to Interactive Decision Making- Elicitation of opinions by means of possibilistic sequences of questions- Searching fuzzy concepts in a natural language data base- Reconfigurable network architecture for distributed problem solving- 4 Algorithms and Software for Interactive Fuzzy Optimization- Interactive decision making for multiobjective linear programming problems with fuzzy parameters based on a solution concept incorporating fuzzy goals- FULP - a PC-supported procedure for solving multicriteria linear programming problems with fuzzy data- 'FLIP': multiobjective fuzzy linear programming software with graphical facilities- FPLP - a package for fuzzy and parametric linear programming problems- An expert system for the solution of fuzzy linear programming problems