Study on Modeling and Application of a Chaotic Series Forecasting Model

Because Chaotic series comprise a large amount of singular points and extreme values, it is of great difficulty to establish a forecasting model to realize accurate prediction of the time series. This paper describes a kind of series characteristic of the given chaotic series information through fractal technology, also describes its structure with Hurst index, calculates out the predictable time, then reconstructs the series disassembled from wavelet transform through phase space, sets up accurate prediction model with the method of the fuzzy neural network, and achieves the application successfully