A neural-fuzzy logic approach for modeling and control of nonlinear systems

Neural networks and fuzzy logic systems are two of the most important results of the research in the area of soft computing. They have been effectively applied to a wide range of applications [1],[2]. While neural networks and fuzzy logic have added a new dimension to many engineering fields of study, their weaknesses have not been overlooked, in many applications the training of a neural network requires a large amount of iterative calculations. Some times the network cannot adequately learn the desired function. Fuzzy systems, on the other hand, are easy to understand because they mimic human thinking and acquire their knowledge from an expert who encodes his knowledge in a series of if/then rules. The problem arises when systems have many inputs and outputs. Obtaining a rule base for large systems is difficult, if not impossible. Promoted by the weaknesses inherited in the technologies and their complementary strengths, the researchers have looked at ways of combining neural networks and fuzzy logic systems. Due to the relative youth of this field of study, a consensus on the best way to utilize their individual strengths and compensate for their individual shortcomings has not yet been established. Consequently, research into neuro-fuzzy systems branches in many directions. The technique used in this work replaces the rule-base of a traditional fuzzy logic system with backpropagation neural network. In this paper we propose an adaptive neuro-fuzzy logic control scheme (ANFLC) based on the neural network learning capability and the fuzzy logic modeling ability. The development of this system will be carried out in two phases: The first phase involves training a multi-layered neuro-mulator (NE) for the forward dynamics of the plant to be controlled; the second phase involves on-line learning of the neuro-fuzzy logic controller (NFLC). Extensive simulation studies of nonlinear dynamic systems are carried out to illustrate the effectiveness and applicability of the proposed scheme.