| Popis: |
The support vector machine, a novel artificial intelligence-based approach developed from statistical learning theory, is used in this work to develop a real-time stage forecasting model. The orders of the input variables are determined by applying the hydrological concept of the time of response, and a two-step grid search method is used to find the optimal parameters, and thus overcome the difficulties of constructing the learning machine. Two structures of models used to perform multiple-hour-ahead stage forecasts are proposed. Validation results from flood events demonstrate that the proposed model can accurately forecast the flood stages one to four hours ahead. Moreover, two statistical tests are used to analyze the forecasting errors. |
