Author(s): Wang Tao; Yang Kai-Lin; Fu Hui; Guo Yongxin; Guo Xinglei
Linked Author(s):
Keywords: South-to-North Water Diversion; Middle Route Project; Water diversion in winter; BP Neural Networks; Levenberg-Marquart; Date; Forecast; Negative air temperature
Abstract: As being building South-to-North Water Diversion Middle Route Project, it is difficult for Ice Forecast because there are no the existing meteorological and hydrological data, the artificial neural networks with the merits as the robustness and fault toleration can find satisfied solutions. By analyzing the meteorological and hydrological character, an ANN model based on feed-forward back-propagation and improved by Levenberg-Marquardt algorithm is applied to forecast dates when the air temperature goes below zero degree in South-to-North Water Diversion Middle Route Project. The forecast results in the winter of 2003-2006 are mostly in good agreement with the measured ones. The accuracy is improved obviously when the midwinter is as the datum points of the statistic date, this result show that the traditional Solar Terms of Chinese Calender is very important for the forecast correlated the air temperature and Climate including the ice condition.
Year: 2010