A Novel Approach to Fuzzy Model Identification Based on Bat Algorithm

Abstract

The identification of a fuzzy model is a complex and nonlinear problem. This can be formulated as a search and optimisation problem and many computing approaches are available in the literature to solve this problem. This research paper is focused on using a new nature inspired approach for fuzzy modeling based on Bat Algorithm which is derived from the behaviour of micro-bats to search for their prey. The bat algorithm approach has been implemented and validated successfully on a rapid battery charger fuzzy controller problem. Currently, the key requirement is real-time solutions to complex problems at a blazing speed. Bat algorithm evolved the optimised fuzzy model within a few seconds as compared to other approaches.