论文部分内容阅读
为了提高径流预测的精度,采用EEMD将非线性、非平稳的径流时间序列分解为若干固有模态分量和趋势项分量,高频分量采用GA-SVM模型进行预测,低频分量采用GA-BP模型进行预测,趋势项采用RBF模型进行预测,然后对各分量进行重构,从而建立了EEMD组合预测模型,并应用于黄河上游主要来水区年来水量预测。结果表明:黄河上游主要来水区年来水量预测误差小于20%的预报合格率为100%,预测精度高,具有较高的实用价值。
In order to improve the accuracy of runoff prediction, EEMD is used to decompose the non-linear and non-stationary runoff time series into a number of intrinsic modes and trend components. The high frequency components are predicted by GA-SVM model, and the low frequency components are estimated by GA-BP model Forecasting, the trend items are predicted by RBF model, and then the components are reconstructed to establish the EEMD combined forecasting model, which is applied to the annual water volume prediction of the main water areas in the upper reaches of the Yellow River. The results show that the forecasting rate of predicting the annual water loss in the main reaches of the Yellow River is less than 20% and the forecasting rate is 100%. The prediction accuracy is high and it has high practical value.