模糊支持向量机在径流预测中的应用

在分析现有径流预测模型局限性的基础上，考虑径流量随时间变化的不确定因素，建立了基于模糊支持向昔机的径流预测模型，使得较近时间的径流数据与较远时间的径流数据相比，对预测精度的提高影响更大．将该模型应用于新疆伊犁河雅码渡站年径流预测中，与传统的支持向量机预测模型比较表明，模糊支持向量机预测精度有较大的提高，并能进一步在其他流域径流预测中推广应用．

Application of fuzzy support vector machine to runoff forecast

The disadvantages of present runoff forecast models are analyzed; and considered the uncertain factor of runoff varied with time, a runoff forecast model is proposed based on fuzzy support vector machine. Comparing with the earlier runoff data, the later ones are more important to improve the prediction accuracy in this model. The model is applied to the annual runoff forecast of Heya Madu Station on Yili River in Xinjiang Autonomous Region and compared with the traditional support vector machine model. The comparison and analysis of the results show that the fuzzy support vector machine model has a better effect than the traditional support vector machine model; and could be further generalized and applied to the runoff forecast of other watershed.