基于EEMD-SVM模型的北洛河上游径流预测

目的]提高北洛河上游径流预报精度，为流域管理及水资源合理配置提供依据。方法]以1971—2014年北洛河上游吴旗水文站实测径流资料为基础，采用EEMD-SVM耦合模型对吴旗站月径流序列进行了模拟预测，并与EEMD-ARIMA和EEMD-NAR两种耦合模型的预测结果进行对比。结果]EEMD-SVM模型的平均绝对误差(MAE)和均方根误差(RMSE)最低，决定系数(R²)和纳什系数(NSE)最高。其中，相比于EEMD-ARIMA和EEMD-NAR模型，EEMD-SVM模型的决定系数(R²)分别提高了186.63%,49.49%。结论]EEMD-SVM模型具有更高的预测精度和更强的非线性拟合能力，可以成功地应用于北洛河上游的月径流预报。同时，研究表明EEMD-NAR模型的预测性能高于EEMD-ARIMA性能。

Runoff Prediction in the Upper Reaches of Beiluo River Based on EEMD-SVM Model

Objective] The aim of this study is to improve the accuracy of runoff forecast in the upper reaches of Beiluo River and provide basis for watershed management and rational allocation of water resources. Methods] Based on the measured runoff data of Wuqi hydrological station in the upper reaches of Beiluo River from 1971 to 2014, the monthly runoff series was predicted by using EEMD-SVM coupling model, and compared with the prediction results of EEMD-ARIMA and EEMD-NAR models. Results] EEMD-SVM model has the lowest mean absolute error and root mean square error, and the highest coefficient of determination(R²)and Nash coefficient. Compared with EEMD-ARIMA and EEMD-NAR models, the R² value of EEMD-SVM model increases by 186.63% and 49.49%, respectively. Conclusion] EEMD-SVM model has higher prediction accuracy and stronger nonlinear fitting ability, and can be successfully applied to the monthly runoff prediction in the upper reaches of Beiluo River. At the same time, the prediction performance of EEMD-NAR model is higher than that of EEMD-ARIMA model.