Study on the Medium and Long Term Runoff Forecast Model Based on Wavelet Analysis
WU Peng-lin · Yellow River · 2013
Aiming at resolving the problem of low accuracy of medium and long term runoff forecast at present,this paper put forward a time series forecast model based on wavelet analysis.Using the wavelet function db(5) to decompose the runoff series at scale 3,the low frequency and high frequency signals are obtained.Then all levels of decomposition signals are predicted with ARMA(2,1) model and their results are compounded to get the predictive value of the original runoff,which compared with the forecast value calculated directly with the ARMA model of annual runoff.The result shows that the former model predicts more accurate and steady than the latter,which verifies the effectiveness of the medium and long term runoff forecast model.