Proceedings of North American Fuzzy Information Processing

Fuzzy logic is presently used to synthesize single layer nonlinear control systems in an increasing range of application areas. The proper role for fuzzy logic is within the framework of a hierarchically structured control system that may incorporate conventional control strategies. The paper focuses on integration of these notions within an analytically tractable framework whereby certain fundame... View full abstract»

This paper introduces a new neuro-fuzzy system, an effective optimization method through a genetic algorithm, a performance criterion for model selection, and a numerical example to illustrate the proposed modeling and control approach. The neuro-fuzzy system is based on the polynomial fuzzy neural network architecture. A new performance criterion is defined based on the Group Method of Data Handl... View full abstract»

We discuss the problem of synthesis of stable Takagi-Sugeno-Kang (TSK) fuzzy systems (1985). This approach is based on a polytopic input-output representation of the fuzzy system model and the fuzzy controller. A procedure for designing fuzzy controllers is also developed. View full abstract»

Summary form only given. The field of the so-called emerging technologies, ie., fuzzy logic control, neuro-fuzzy control, has witnessed quite a sizable introduction of application products, mainly from Japan and Europe and recently from the US. The soft-computing community (fuzzy logic, neural networks and genetic algorithms) has begun to respond to the barrage of concerns about the analytical iss... View full abstract»

Humans have excelled at the complex perceptual tasks involved in visual interpretation. Progress has been consistent but fairly slow in performing vision related problems by computer. In an effort to bring more human-like processing into computer vision activities, rule-based systems have been utilized. The uncertainty and vagueness present in image analysis suggest fuzzy logic is a natural paradi... View full abstract»

Fuzzy clustering algorithms have been used in the automatic segmentation of large multi-dimensional data sets such as 2D magnetic resonance images. Although the clustering algorithms perform well, they are very time consuming. Better performance at a clustering stage is achieved if the data-set can be partially classified before clustering is applied. Further, if the pre-classification is itself f... View full abstract»

This paper presents a new mechanism for determining feature correspondences for object recognition, based on fuzzy set theory. The new method applies unary and binary constraints from the model, taking uncertainty characteristics of the measurement process into consideration. Experiments with both unoccluded and occluded images show that the method selects an appropriate set of correspondences, es... View full abstract»

A task-oriented algorithm that selectively processes image data is developed. The task of the algorithm is detecting/locating circular shaped road signs. Structural characteristics of the road scenes allow us to extract high level (contextual) information directly from the scene. This information is used as evidence for a road sign and aggregated with more evidence obtained by low level (pixel bas... View full abstract»

In this paper, the relationship between the existing axioms and the corresponding CNF and DNF based on the n-variable-m-dimensional fundamental clauses and phrases is discussed. The AND (OR) map is the general framework of the fuzzy map and the Karnaugh map without losing any primary combined concepts. In addition, the general algorithm to make the n-variable-m-dimensional CNF (DNF) from a DNF (CN... View full abstract»

The main aim of this paper is to point out that a connective rule should be understood as a consistent family of connectives, in such a way that given any finite sequence of values we can evaluate its connective value. A connective rule is what we really need in practice, not a single connective operator. Only in some few cases we can characterize such a connective rule by means of a unique (assoc... View full abstract»

A generalization of the t-norm and t-conorm called the uni-norm is defined. These operators allow for an identity element lying anywhere in the unit interval, rather than at one or zero as in the case of t-norms and t-conorms respectively. Some properties of these uni-norms are discussed. Two particular families of uni-norms (R* and R/sub */) are introduced. Weighted aggregations using these opera... View full abstract»

Fuzzy (logic) neurons have triangular norms as synaptic operators and triangular s-norms as somatic operators. MIN and MAX are often the chosen t-norm/s-norm pair. This choice is mainly due to the simplicity of the calculations and to the fact that a layer of MAX-MIN neurons implements the widely-used MAX-MIN composition. From a neural networks perspective (i.e. having in mind the ability to use t... View full abstract»

A fuzzy-valued measure of the degree of matching of two fuzzy sets is proposed based on the Jaccard index applied to the level sets. This measure satisfies a generalization the criteria of Dubois and Prade (1982) for point-valued matching indices. A similarity index is obtained by combining the matching index with a proximity assessment. These fuzzy-valued indices provide a more comprehensive summ... View full abstract»

Decision analysis and decision support are an area in which applications of fuzzy set theory have been found since the early 1970s. Still, there are areas, such as fuzzy control, that have gained much wider acceptance in practice than fuzzy decision analysis. This paper describes where fuzzy decision support stands now, and what would have to be done for it to gain wider acceptance. One of the maj... View full abstract»

Discusses the framework for fuzzy criterion set-based dynamic programming, which generalizes the tools that are applicable to an array of fuzzy decision models, particularly those represented by the generalized reservoir operation and the stochastic inventory models. The objective is to maximize the expected fuzzy criterion function of the product of fuzzy criterion sets. We outline certain proper... View full abstract»

Deals with a multi-stage decision process with fuzzy transitions, which is termed a 'fuzzy decision process'. We consider such a process where both states and actions are assumed to be fuzzy, from the point of view of a dynamic fuzzy system which has been developed by the authors. A time-average fuzzy reward for the process is described by a fuzzy number; and a partial order of convex fuzzy number... View full abstract»

Proposes a method to solve qualitative locational choice problems using fuzzy decision tables as a matching model. Firstly, the technique of crisp decision tables is explained, and their use in locational choice problems is advocated. Subsequently, fuzzy extensions of decision tables are defined. Next, using a brief example, it is shown that fuzzy decision tables can be used efficiently for evalua... View full abstract»

Presents the results of applying the Levenberg-Marquardt method (K. Levenberg, 1944, and D.W. Marquardt, 1963), which is a popular nonlinear least-squares method, to the ANFIS (Adaptive Neuro-Fuzzy Inference System) architecture proposed by Jang (IEEE Trans. on Systems, Man and Cybernctics, vol. 23, no. 3, pp 665-685, May 1993). Through empirical studies, we discuss the strengths and weaknesses of... View full abstract»

Hybrid regression analysis is proposed for modeling randomness and fuzziness in a regression model. For regression analysis involving fuzzy numbers, weighted fuzzy arithmetic is defined and used as a replacement for conventional fuzzy arithmetic. A method of hybrid least-squares linear regression is developed. Reliability measures for hybrid regression are also defined. A numerical example of hybr... View full abstract»

This paper presents a new method for the stability analysis of fuzzy control systems using Petri nets. The proposed method can describe the fuzzy control system by using the next-state function. The dynamical behavior of the system is able to be understood easily. A simulation is carried out in order to demonstrate the proposed stability analysis method. View full abstract»

Stability analysis using the hyperstability approach is proposed for a certain class of fuzzy controllers. Frequency conditions for PI and PD-like fuzzy controllers are established both in continuous-time and discrete-time cases. A robustness criterion is also given. View full abstract»

The elevator group control systems (EGCS) are the control systems that manage systematically three or more elevators in order to efficiently transport the passengers. Most of the EGCS's have used the hall call assignment method to assign elevators in response to passenger's calls. This paper proposes a control strategy generation method, a hall call assignment method based on fuzzy theory and the ... View full abstract»

A fuzzy control system has been developed and tested for a continuous, hot-air drying process. The system has been tested for soybean drying, however many elements are common to other food processing control problems. The control system includes: a supervisory level which considers quality criteria (fuzzy and crisp); a feedforward controller which sets process residence time; and a feedback correc... View full abstract»

This paper is devoted to a new concept for observing the states of nonlinear plants. It is based on ready made fuzzy controllers. The concept is applied to an inverted pendulum system, and excellent simulation results are obtained. The issue of parameter uncertainties is also discussed, and robustness behaviour of the observer is shown. View full abstract»