Abstract:In this paper, we introduce and investigate a new category of fuzzy inference systems based on information granulation and genetic optimization used to system identification. We show the applications of such systems to identification of nonlinear systems. The formal framework of information granulation and resulting information granules themselves become an important design facet of the fuzzy models. By embracing fuzzy sets, the model is geared towards capturing essential relationship between information granules rather than concentrating on plain numeric data. Information granulation realized with the use of the commonly exploited C-Means clustering helps determine the initial values of the parameters of the fuzzy models. This in particular concerns such essential components of the rules as the initial apexes of the membership functions standing in the premise part of the fuzzy rules and the points of the polynomial functions standing in the consequence part. The initial apexes (center points) of the membership functions based on C-Means algorithm are tuned with the aid of the genetic algorithm (GA), while the tuned apexes are also used to adjust the points of the consequent polynomials (conclusions) of the rules. In particular, the initial apexes of the membership functions and the initial points of the consequent polynomials are adjusted and updated every time through successive evolution process. The overall design methodology involves a hybrid structural and parametric optimization. Genetic algorithms and C-Means clustering are used to optimize the model with respect to its structure and parameters. To determine the structure and estimate the values of the parameters of the fuzzy model we consider the successive tuning method with generation-based evolution by means of genetic algorithms. The model is evaluated with the use of numerical experimentation and its quality is compared with respect to some other fuzzy models already encountered in the literature.

Garel - A Hybrid Genetic Learning in Fuzzy Relational Equations

Fuzzy Relational Structures: Learning Alternatives For Fuzzy Modeling

Genetic Learning of Fuzzy Cognitive Maps

Using Genetic Algorithm To Design Hierarchical Flc System

A Genetic Approach to Modeling Fuzzy Systems Based on Information Granulation and Successive Generation-Based Evolution Method.

A genetic algorithm for rule extraction in fuzzy adaptive learning control networks

A hybrid coevolutionary algorithm for designing fuzzy classifiers

A Genetic-Based Method for Training Fuzzy Systems

A Hybrid Approach for Automatic Generation of Fuzzy Inference Systems Without Supervised Learning

Learning fuzzy rules for modeling complex classification systems using genetic algorithms

Design of Fuzzy Relation-Based Polynomial Neural Networks Using Information Granulation and Symbolic Gene Type Genetic Algorithms

Learning and Tuning Fuzzy Logic Controllers Through Genetic Algorithm

Genetic-Programming-Based Architecture of Fuzzy Modeling: Towards Coping with High-Dimensional Data

GARA: A novel approach to Improve Genetic Algorithms' Accuracy and Efficiency by Utilizing Relationships among Genes

Adaptive Fuzzy Logic Controllers Using Hybrid Genetic Algorithms

Fuzzy Reasoning Based on Genetic Algorithms

Identification of Fuzzy Models with the Aid of Evolutionary Data Granulation

Learning Fuzzy Cognitive Maps Using a Genetic Algorithm with Decision-Making Trial and Evaluation

Fuzzy Relational Equations: Bridging Theory, Methodology And Practice

Identification of fuzzy relation models using hierarchical fair competition-based parallel genetic algorithms and information granulation

Cascade Architectures of Fuzzy Neural Networks