A study of the hydrological prediction model based on wavelet de-noise method
Geng Qu, Min Feng-yang, Xiaohu Guo, Yonghui Zhu · 2010 Sixth International Conference on Natural Computation · 2010
The general hydrological time series data inevitably contained “noise” for various factors. The existence of “noise” changed the autocorrelation and covered the real change characteristics of the hydrological time series. We based this study on the stochastic model of wavelet de-noise and applied the model to the hydrological prediction. In this paper five decades runoff data of Yichang Station (1952-2002, monthly) were analyzed and were de-noised by the wavelet method. A stochastic model is established on the basis of the variable feature of de-noised series data and the hydrological time series data were predicted by the model. Results of the study indicate that the wavelet analysis has great priority in de-noise procession and trend prediction. The predicted model on the basis of the de-noise time-series has a high precision for the hydrological data prediction.