Many complicated systems can not be defined by conventional methods which haven't been built effective mathematic model and control method. Recently, both fuzzy and neural network technology has become a very active branch in intelligence control theory, wavelet analysis technology with local performance of time-frequency and multiresolution function has been rapidly developed in the past several years. So, based on the combination of fuzzy logical technique,wavelet analysis and neural network, this paper researches a new model, called fuzzy wavelet neural network. The main research works in this article are as follows:First, it offers an overview of the methods of identification and control for nonline ar systems and their current situation,and a survey to research progress on the combina tion of fuzzy logical technique,wavelet analysis and neural network at present is made.Second, dased on the combination of T-S fuzzy model and the feedforward neural networks, this paper researches a new model, called fuzzy neural network, and research a mixed learning algorithm based on the combination of gradient descent algorithm andα? LMS algorithm. Here, The nonlinear parameters are learned by gradient descent al- gorithm and linear parameters,viz.,weights are learned byα? LMS algorithm.The app roximation capabilities of the model to nonlinear function are theoretically proved.Finally,by utilizing two important properties,viz.,multi-resolution and compression of wavelets in conjunction with fuzzy neural networks, two new fuzzy wavelet neural networks are proposed here for non-linear system identification. The simulations to the static system demonstrate that the proposed two fuzzy wavelet neural networks are optimal. |