Study On Method Of Chaotic Time Series Forecast And Its Application In Market Requirement

Society economic system nonlinearity nature decides the way of building nonlinear model to forecast. As a nonlinear forecasting method, chaotic time series forecasting method is on the basis of chaos theory, which breakthrough the limitation of build subjective model. It has successfully applied in many fields by analyzing the inner rule of time series for forecasting.The thesis applies chaotic time series method in demand forecasting creatively. By phase space reconstruction, choosing the most suitable delay time and embedding dimension in order to embed time series which reflect the demanding into the phase space. Based on the judgment of chaotic system by calculation of Lyapunov exponent, adding weight zero-rank weight local method and adding weight one-rank local region method are used for forecasting. The thesis works out programs of C-C method, G-P arithmetic, adding weight zero-rank local method and adding weight one-rank local region method. Forecast result indicate that comparing with traditional forecasting method, chaotic time series method can improve the precision and reliability of forecast result.