Optimizing fuzzy neural network controller based on NSGA-II

Abstract

Fuzzy system is well known system for its capabilities by solving various kinds of control problems. In this article, we proposed a method for optimizing the fuzzy logic controller through genetic algorithms and neural networks. The optimization method is composed of (1) the neural network with clustering is designed to learn an initial rule base, if no prior knowledge about the system is available. Then (2) the learned rule base is optimized through Non Dominated Sorting Genetic Algorithm-II (NSGA-II). To measure the validity and efficiency of the proposed system these optimized rules is applied to control a truck like robot system. Experimental results demonstrate better performances of proposed Hybrid Neural-Genetic-Fuzzy System (NGFS) than other existing systems.