Study Of Fuzzy And Neural Network

The rapid development of modern technology makes control theory develop in the direction of more complication and more accuracy than before. Fuzzy control and neural networks control attract more and more people just because they have the charactors of special noneline and need not establish the mathematical model. Fuzzy system is good at expressing knowledge and its logical reasoning is similar to man' s thought. But this system depends on the man' s subjective factor too much and lack the capacity of adaptation and learning. Neural network has a Variable construction. To the most importace it has the characters of self-organization and self-learning. But its network parameters lack the physics meaning and easily trap in the local convergence at the same time. It is an inevitable tendency to combine these two systems in order to absorb their advantages .In this paper, the theories of Fuzzy logic control, Neural networks, and Fuzzy and Neural networks are introduced. Noticing the difficulties in optimizing the input space and withdrawing fuzzy rules mentioned in the dissertation, we put clustering algorithm into fuzzy neural networks and withdraw system' s character, optimize the input space in this method in order to set up a self-adaptive model. The effectiveness of the method is veritfied through simulationt tests. The main contribution of this dissertalion is summatized as aollows.1. To introduce the knowledge of fuzzy control, neural netword, fuzzy and neural network and clustering algorithm.2. T-S model is a novel fuzzy reasoning model which replaces the parameters of traditional resoning system with linear partial equation. Thereore, it can generate complicated nonlinear equation with fewer fuzzy rules. In the paper,there are two fuzzy and neural networks system based on T-S fuzzy reasoning model. We use the clustering algorithm method to get the number of the fuzzy rules and optimize the input space. The conclusive section and the adaptability of rules of the fist system are both constructed by neural networks. The second system is the simple struction of the first one. It uses the fuzzy C clustering algorithm to put the adaptive degree of the fuzzy rules and simplify the struction of the neural network of conclusion.Both of the two systems combine the nonsupervised algorithm and the gradient based algorithm, we adjust the distributing condition of the clustering point according to the density of distribution of data.3. As the ANFIS system is a flexible system, and MATLAB has powerfulfunction of calculation and simulation, we construct an ANFIS Fuzzy and neural network model based on sub clustering algorithm , and offer the instance based on MATLAB. This model can get the number of the rules,confirm system construction and initialize original parameters according to clustering result automatically. The result parameters can be adjusted by algorithm.In the process we can draw the conclusion that the new algorithm is better than the old one in adaptation, computational complexity and the modeling accurcy.