Wavelet neural network and application

Wavelet analyzing has localized characteristic in time and frequency, is that a kind of function analytical method including yardstick flexible and Translation. Wavelet Neural Network is a kind of new-type neural network developed on the basis of Wavelet analyzing in recent years, it utilizes the wavelet base as neuron function to produce the network, have advantages of simple algorithm and quickly restrain the speed, and apply to extensively the approximation of non-linear function, system control and pattern recognition, etc. This paper summarize Wavelet Neural Network principle and basic model, combine instance introduce wavelet neural network typical application, discuss three kind of WaveletNeural Network (continuous wavelet transformation, wavelet transformation based on block frame, orthogonal wavelet transformation)characteristic on this basis, analyse wavelet analyzing main defect that exist at present, and look forward to the development of Wavelet Neural Network.