SSA分解预校正模型在辽宁西部水库年径流预测中的应用研究

Application analysis of SSA Decomposition and Pre-correction Model in the Prediction of Annual Runoff of Reservoirs in Western Liaoning Province

本文采用SSA分解预测模型结合灰色预测模型对辽宁西部某水库年径流进行预测。研究结果表明:SSA分解预测模型可降低年径流低频振荡的显著性,可对灰色预测模型预测的年径流系列进行预先校正,提供模型预测精度。相比于校正前的灰色模型,校正后的模拟预测年径流系列和水库实测年径流系列相对误差减少15.2%,年径流相关系数提高0.21,研究成果对于水库年径流预测方法提供参考价值。

In this paper,the prediction of annual runoff of a reservoir in the west of Liaoning Province is car- tied out by using SSA decomposition prediction model and gray prediction model.The results show that SSA decomposing prediction model can reduce the significance of annum runoff low frequency oscillation, and can predict the runoff series predicted by gray forecast model in advance,which can provide model prediction ac- curacy.Compared with the gray model before correction, the relative error of the calibrated annual runoff series and the annum runoff series of reservoirs are reduced by 15.2% and the annual runoff correlation coefficient is increased by 0.21.The research results can Keywords provide reference for the prediction of annual runoff.