Wave-Ann Model In The Application Of Energy Forecast

In the past national economic depended on local resources had rapid developed. Butthe poor domestic per capita amount of resources, low energy efficiency and extensivedevelopment structure led directly to increase the main mineral resources of foreigndependence, which not only increases the risk of energy in our country, causes the veryserious harm to the environment, but also causes threat for economic and socialsustainable development. According to related field energy forecast, it increase theenergy utilization rate, and have the important meaning for the warning of the risks ofenergy, protecting the ecological environment, and protecting the national economyproperty.This paper analyses the energy situation, development trend, the role the energyforecast had in the energy industry in China, and now our country’s energy forecastresearch situation. According to energy forecast problems and puts forward use theWave-Ann model to forecast energy.This paper briefly introduces the theory knowledge of wavelet analysis, compare thewavelet analysis with the Fourier transform and window Fourier transform, andexpounds the powerful ability in the wavelet analysis in time-frequency local processing,and the Mallat rapid method in the practical application of the discrete wavelet transformmethod, and introduces the artificial neural network model to provide theoretical basisfor the Wave-Ann model.Wavelet analysis has localization time-frequency analysis ability, can use the"convergence" to realize time and frequency signal analysis, separate the effective signalfrequency components, realize the effective signal information extraction. Artificialneural network has excellent since learning and adaptive ability, good nonlinear fittingability. The combined forecasting model (Wave-Ann model) which based on the waveletanalysis and artificial neural network have combined with the advantages of both,effectively identify the main frequency signal composition and partial information, toavoid the input variable data, slow convergence speed or other weakness, and using artificial neural network to realize the history and future of the nonlinear sequence datafitting, improve the prediction precision and predict speed.Based on the Wave-Ann model, the article studies the Wave-Ann model in theapplication of energy forecast fields, such as the mid-to-long-term hydrologicalforecasting areas and the wind power forecasting areas, this provides theory basis forhydroelectric power dispatching and wind power grid dispatching.