The Study Of Cooperation Fuzzy Chaotic Neural Network

Although artificial intelligence (AI) has been developed for many decades and lots of achievements have been obtained, researches on the relations between brain, thinking and computing are started just now, study on the computing principle of brain and its complexity,on the mechanic of learning- association-memory and its imitation are in its beginning. In order to better understand the process of information processing in the brain, it is necessary that we imitate some factions of brain with artificial method, such as fuzzy logical (FL), artificial neural networks (ANN),chaos and so on, which are all now edge subjects. After the development of various edge subjects, researches are always focused on the relations between them, and find out their common essence, so do researches on FL, ANN, and chaos.FL, ANN, and chaos dynamics all are able to reflect one aspect of information processing in brain, however, the existing ANNs can only reflect one or two aspects, this is, they fail to establish new method to realize AI from three aspects at the same time, A new method which is based on FL, ANN and chaos is proposed in this paper, so we tried to fully understand of information processes in brain from another aspect so as to push AI forward.A fuzzy Chaos neural network models named collaboration fuzzy Chaos neural network was designed in this paper. It established the model's output mathematic function . learning algorithm and analyzed the model's dynamics characteristics The model was structurally neural network with chaotic features and functionally fuzzy system. And modelling based on the network was studied.Main works of the dissertation are as follows:Firstly, introduced the model of sigmoid CNN, and its dynamics characteristics, such as Lyapunov exponent (LE), bifurcation and dissipative have been analyzed thoroughly.In order to better imitate the structure and the function of human brain,...