AIRCC's International Journal of Computer Science and Information Technology

Pintu Chandra Shill ¹, Bishnu Sarker ², Kazuyuki Murase ¹

Affiliations
1 Department of System Design Engineering, University of Fukui, 3-9-1 Bunkyo, Fukui 910-8507, JP
2 Department of Computer Science and Engineering, KUET, Khulna 9203, BD

Abstract

Fuzzy systems generally works based on expert knowledge base. Fuzzy Expert knowledge base derived from the heuristic knowledge of experts or experience operators in the form of fuzzy control rules and membership functions (MFs). The major difficulties for designing a fuzzy models and controllers are identify the optimized fuzzy rules and their corresponding shape, type and distribution of MFs. Moreover, the numbers of fuzzy control rules increases exponentially with the number of input output variables related to the control system. For this reason it is very difficult and time consuming for an expert to identify the complete rule set and shape of MFs for a complex control system having large number of input and output variables. In this paper, we propose a method called evolutionary fuzzy system for tuning the parameters of fuzzy rules and adjust the shape of MFs through evolutionary algorithms in order to design a suitable and flexible fuzzy models and controller for complex systems. This paper also presents new flexible encoding method methods for evolutionary algorithms. In evolutionary fuzzy system, the evolutionary algorithms is adapted in two different ways. Firstly, generating the optimal fuzzy rule sets including the number of rules inside it and secondly, selecting the optimum shape and distribution of MFs for the fuzzy control rules. In order to evaluate the validity and performance of the proposed approach we have designed a test strategy for the modeling and control of nonlinear systems. The simulation results show the effectiveness of our method and give better performance than existing fuzzy expert systems.

Keywords

Fuzzy Expert System, Optimization, Evolutionary Algorithms (EAs), Evolutionary Fuzzy System and Nonlinear System.

Full Text

   

Md. Amjad Hossain ¹, Pintu Chandra Shill ², Bishnu Sarker ¹, Kazuyuki Murase ²

Affiliations
1 Department of Computer Science and Engineering, KUET, Khulna 9203, BD
2 Department of System Design Engineering, University of Fukui, 3-9-1 Bunkyo, Fukui 910-8507, JP
3 Department of Computer science and Engineering, KUET, Khulna 9203, BD

Abstract

Fuzzy rule based models have a capability to approximate any continuous function to any degree of accuracy on a compact domain. The majority of FLC design process relies on heuristic knowledge of experience operators. In order to make the design process automatic we present a genetic approach to learn fuzzy rules as well as membership function parameters. Moreover, several statistical information criteria such as the Akaike information criterion (AIC), the Bhansali-Downham information criterion (BDIC), and the Schwarz-Rissanen information criterion (SRIC) are used to construct optimal fuzzy models by reducing fuzzy rules. A genetic scheme is used to design Takagi-Sugeno-Kang (TSK) model for identification of the antecedent rule parameters and the identification of the consequent parameters. Computer simulations are presented confirming the performance of the constructed fuzzy logic controller.

Keywords

Genetic Algorithms (GAs), Fuzzy Logic Controller (FLC), Statistical Information Criteria, Singular Value Decomposition (SVD), Takagi-Sugeno-Kang (TSK) Model.

Full Text