Volume 6     Volume 5     Volume 4     Volume 3     Volume 2     Volume 1

Issue 4     Issue 3     Issue 2     Issue 1

How to design a discrete supervisory controller for real-time fuzzy control systems
N. Muskinja, B. Tovornik, D. Donlagic
Fac. of Electr. & Comput. Sci., Maribor Univ., Slovenia
vol. 5, no. 2, pp. 161-6, May 1997

The task of the supervisory controller is to stabilize the systems states within a bounded region defined by designer. In this paper, a discrete approach to the solution of the stable fuzzy control system is presented. It is proved that the fuzzy control system equipped with the discrete supervisory controller is globally stable in the Lyapunov sense. Finally, a fuzzy controller with a discrete supervisory controller is applied to the balance control system, both in simulations and in the real-time implementation.

Adaptive fuzzy control: experiments and comparative analyses
R. Ordonez, J. Zumberge, J. T. Spooner, K. M. Passino
Dept. of Electr. Eng., Ohio State Univ., Columbus, OH, USA
vol. 5, no. 2, pp. 167-88, May 1997

Advances in nonlinear control theory have provided the mathematical foundations necessary to establish conditions for stability of several types of adaptive fuzzy controllers. However, very few, if any, of these techniques have been compared to conventional adaptive or nonadaptive nonlinear controllers or tested beyond simulation; therefore, many of them remain as purely theoretical developments whose practical value is difficult to ascertain. In this paper we develop three case studies where we perform a comparative analysis between the adaptive fuzzy techniques in Spooner and Passino (1995,1996) and some conventional adaptive and nonadaptive nonlinear control techniques. In each case, the analysis is performed both in simulation and in implementation, in order to show practical examples of how the performance of these controllers compares to conventional controllers in real systems.

New triangular operator generators for fuzzy systems
S. Roychowdhury
Dept. of Math. & Comput. Sci., Tulsa Univ., OK, USA
vol. 5, no. 2, pp. 189-98, May 1997

Triangular operators (t-operators) form an integral part in the design and analysis of fuzzy systems. Simple monotonic, continuous, nonconditional functions are used in an operator generator to generate t-operators. Depending on the operator generator and the function that it uses, it becomes easier to characterize and classify the families of t-operators. In this paper, the author proposes two operator generators that will extend the domain of triangular operators in the realm of fuzzy set theory. The conventional operator generators generate a t-norm and a t-conorm by using a decreasing function and an increasing function, respectively. In contrast, in this study, increasing functions generate t-norms, while decreasing functions generate t-conorms, respectively.

Dynamic non-Singleton fuzzy logic systems for nonlinear modeling
G. C. Mouzouris, J. M. Mendel
Dept. of Electr. Eng. Syst., Univ. of Southern California, Los Angeles, CA, USA
vol. 5, no. 2, pp. 199-208, May 1997

We investigate dynamic versions of fuzzy logic systems (FLSs) and, specifically, their non-Singleton generalizations (NSFLSs), and derive a dynamic learning algorithm to train the system parameters. The history-sensitive output of the dynamic systems gives them a significant advantage over static systems in modeling processes of unknown order. This is illustrated through an example in nonlinear dynamic system identification. Since dynamic NSFLS's can be considered to belong to the family of general nonlinear autoregressive moving average (NARMA) models, they are capable of parsimoniously modeling NARMA processes. We study the performance of both dynamic and static FLSs in the predictive modeling of a NARMA process.

Fuzzy neural network in case-based diagnostic system
Qiang Liu Zhi, F. Yan
Dept. of Comput. Sci., Melbourne Univ., Parkville, Vic., Australia
vol. 5, no. 2, pp. 209-22, May 1997

Diagnosing electronic systems for symptoms supplied by customers is often difficult as human descriptions of symptoms are for the most part uncertain and ambiguous. As a result, traditional expert systems are not effective in providing reliable analysis, often require a large set of rules, and lack flexibility in terms of learning and modification. In this paper, we propose a fuzzy logic-based neural network (FLBN) to develop a case-based system for diagnosing symptoms in electronic systems. We demonstrate through data obtained from a real call-log database that the FLBN is able to perform fuzzy AND/OR logic rules and to learn from samples. Such a system is simple to develop and can achieve the performance similar to that of the human expert.

A fuzzy clustering-based rapid prototyping for fuzzy rule-based modeling
M. Delgado, A. F. Gomez-Skarmeta, F. Martin
Dept. de Ciencias de la Comput. e Inteligencia Artificial, Granada Univ., Spain
vol. 5, no. 2, pp. 223-33, May 1997

This paper presents different approaches to the problem of fuzzy rules extraction by using fuzzy clustering as the main tool. Within these approaches we describe six methods that represent different alternatives in the fuzzy modeling process and how they can be integrated with a genetic algorithms. These approaches attempt to obtain a first approximation to the fuzzy rules without any assumption about the structure of the data. Because the main objective is to obtain an approximation, the methods we propose must be as simple as possible, but also, they must have a great approximative capacity and in that way we work directly with fuzzy sets induced in the variables input space. The methods are applied to four examples and the errors obtained are specified in the different cases.

Reconstruction problem and information granularity
G. Bortolan, W. Pedrycz
LADSEB, CNR, Padova, Italy
vol. 5, no. 2, pp. 234-48, May 1997

The paper elaborates on the representation and reconstruction of numerical and nonnumerical data in fuzzy modeling. Proposed are general criteria leading to the distortion-free interfacing mechanisms that help transform information between the systems (or modeling environments) operating at different levels of information granularity. Distinguished are three basic categories of information: numerical, interval-valued, and linguistic (fuzzy). Since all of them are dealt with here, the paper subsumes the current studies concentrated exclusively on representing fuzzy sets through their numerical representatives (prototypes). The algorithmic framework in which the distortion-free interfacing is completed is realized through neural networks. Each category of information is treated separately and gives rise to its own specialized architecture of the neural network. Similarly, these networks require carefully designed training sets that fully capture the specificity of the reconstruction problem. Several carefully selected numerical examples are aimed at the illustration of the key ideas.

Effects of membership function parameters on the performance of a fuzzy signal detector
J. R. Boston
Dept. of Electr. Eng., Pittsburgh Univ., PA, USA
vol. 5, no. 2, pp. 249-55, May 1997

This paper describes a signal-detection algorithm based on fuzzy logic. The detector combines evidence provided by two waveform features and explicitly considers uncertainty in the detection decision. The detector classifies waveforms including a signal, not including a signal, or being uncertain, in which case no conclusion regarding presence or absence of a signal is drawn. Piecewise linear membership functions are used, and a method to describe the membership functions in terms of two parameters is developed. The performance of the detector is compared to a Bayesian maximum likelihood detector, using brainstem auditory evoked potential signals in simulated noise, and the effects of the steepness (slope) and overlap of the membership functions on detector performance are evaluated. By varying the membership function steepness and overlap, the fuzzy detector can almost completely eliminate classification errors at the cost of a large number of uncertain classifications or it can be made to perform similarly to the Bayesian detector.

Rule-based modeling of nonlinear relationships
W. Pedrycz, M. Reformat
Dept. of Electr. & Comput. Eng., Manitoba Univ., Winnipeg, Man., Canada
vol. 5, no. 2, pp. 256-69, May 1997

We discuss a problem of rule-based fuzzy modeling of multiple-input single-output nonlinear relationships f: R/sub n/ to R. The model under investigation is viewed as a collection of conditional statements "if state Omega , then y=g/sub i/(x,at)", i=1,2,...N with Omega /sub i/ being a fuzzy relation defined in the space of the input variables. In contrast to the commonly encountered identification approach, based exclusively upon discrete experimental data, the one proposed in this study is concerned with the rule-based modeling exploiting the available nonlinear input-output relationship. The main thrust is in the development of a relevant fuzzy partition of the input variables. We introduce and study criteria of separability and variability as the key means guiding a distribution and granularity of the linguistic labels forming the condition part of the local models.

Robust clustering methods: a unified view
R. N. Dave, R. Krishnapuram
Dept. of Mech. Eng., New Jersey Inst. of Technol., Newark, NJ, USA
vol. 5, no. 2, pp. 270-93, May 1997

Clustering methods need to be robust if they are to be useful in practice. In this paper, we analyze several popular robust clustering methods and show that they have much in common. We also establish a connection between fuzzy set theory and robust statistics, and point out the similarities between robust clustering methods and statistical methods such as the weighted least-squares technique, the M estimator, the minimum volume ellipsoid algorithm, cooperative robust estimation, minimization of probability of randomness, and the epsilon contamination model. By gleaning the common principles upon which the methods proposed in the literature are based, we arrive at a unified view of robust clustering methods. We define several general concepts that are useful in robust clustering, state the robust clustering problem in terms of the defined concepts, and propose generic algorithms and guidelines for clustering noisy data. We also discuss why the generalized Hough transform is a suboptimal solution to the robust clustering problem.

Fuzzy-controlled perceptual coding of videophone sequences
A. Leone, A. Bellini, R. Guerrieri
Dipartimento di Elettronica, Inf. e Sistemistica, Bologna Univ., Italy
vol. 5, no. 2, pp. 294-303, May 1997

This paper describes a fuzzy quality-controller for video coding systems based on the H.261 recommendation. The picture quality is tuned by a classification of the blocks of each frame of the sequence based on a simplified model of the human visual system. The quantization parameter is evaluated according to this classification making more relevant parts be coded with finer details at the expense of less significant regions. The system is based on a restricted number of fuzzy rules and can be easily reconfigured if a different classification criterion must be adopted. The subjective quality of coded sequences is better than the one obtained by the reference model at the cost of higher computational tasks.

Fuzzy control of multivariable nonlinear servomechanisms with explicit decoupling scheme
J. Nie
Commun. Res. Lab., McMaster Univ., Hamilton, Ont., Canada
vol. 5, no. 2, pp. 304-11, May 1997

This paper deals with the problem of controlling multivariable nonlinear servomechanisms by means of fuzzy approaches. A specific system under consideration is a passive nonlinear line-of-sight (LOS) stabilization system with strong interactions between two channels. By using the concept of decentralized control, a control structure is developed that is composed of two control loops, each of which is associated with a single-variable fuzzy controller and a decoupling unit. A simplified fuzzy control algorithm is used to implement the fuzzy controller. We propose two novel approaches to designing the decoupling units. The first one is based on the theory of fuzzy reasoning whereas the second scheme relies on the principle of adaptation. Extensive simulation studies on the LOS system have demonstrated the feasibility and effectiveness of the proposed approach.