基于支持向量机的中长期入库径流预报

采用基于支持向量机的预测模型对水库中长期入库径流进行预报,建立径流预报的SVM模型。预报因子的优劣决定着预测精度的高低。为了提高预报精度,尝试采用模糊优选法对预报因子进行优选。将所建模型应用于新疆雅马渡站的径流预测中,并与没有进行预报因子优选的SVM模型进行比较。结果表明,进行预报因子优化后的SVM模型明显提高了径流的预报精度,具有更高的应用价值。

Long-term Runoff Forecast Based on the Support Vector Machine

Forecast model based on the support vector machines to forecast the reservoir long-term runoff is used and the SVM runoff forecast model is established. The merits of forecast factors determine the level of forecast accuracy. In order to improve forecast accuracy, the fuzzy optimization method is tried to optimize forecast factors. The model is applied to the runoff forecast of Yamadu Station in Xinjiang, and compared with the SVM model which has not optimized the forecast factors. The results show that the SVM model which has optimized the forecast factors significantly increases the runoff forecast accuracy and has better value.