Research and Application of Basin Rainfall Series Forecasting Based on WA-SVM
Xingyuan Song · China Rural Water and Hydropower · 2007
To enhance the precision of medium-and-long term rainfall forecast,the methods of wavelet analysis and support vector machine were introduced into the fields of hydrology series forecasting.Moreover,the rainfall estimate model was built up based on wavelet analysis and support vector machine(WA-SVM).It decomposes the original complex rainfall series to different layers through wavelet analysis,forecasts each layer by means of SVM,and finally obtains the forecasting results of the original time series by composition.The model is applied to estimate the monthly rainfall series in the watershed,and compare with the results of support vector machine(SVM).The result indicates that the forecast accuracy is elevated obviously.