本文探索了支持向量机在中长期径流预报中的应用。在支持向量机建模过程中引入了径向基核函数,简化了非线性问题的求解过程,并应用SCE-UA算法辨识支持向量机的参数。在SCE-UA搜索过程中进行了指数变换,以快速准确的找到最优参数。与人工神经网络模型预报结果比较显示,该模型能提高径流中长期预报的精度。
The support vector machine (SVM) method is applied to forecast long-term run-off. The radial basic core function is introduced in the establishment of the model describing the run-off hydrograph, and the SCE-UA algorithm is applied to identify the parameters of SVM. The exponential transformation is used to help quickly and precisely search the optimal parameters. The comparison of forecast result between proposed method and artificial neural network (ANN) method indicates that the new method possesses higher forecasting accuracy.