论文部分内容阅读
针对支持向量机在水文过程应用分析中存在的问题,该文将小波变换和支持向量机相结合建立水文时序趋势分析模型。首先对水文序列通过小波变换进行预处理,把处理后序列分解成不同时间尺度下的子序列,然后用支持向量机对各子序列分别进行模拟和预测,将这些支持向量机的预测结果通过小波逆变换重构水文时间序列,建立基于小波变换的支持向量机水文过程趋势分析模型,以三门峡水文站天然月径流时序为例进行应用验证。研究结果表明:与传统的支持向量机、神经网络等预测模型相比,本文模型在预测精度和时间长度上均优于前二者。
Aiming at the problems existing in the application of support vector machine in hydrological process, this paper combines the wavelet transform with support vector machine to establish the hydrological time series trend analysis model. Firstly, the hydrological sequence is preprocessed by wavelet transform, and the processed sequence is decomposed into sub-sequences under different time scales. Then the support vector machine (SVM) is used to simulate and predict each subsequence separately. The prediction results of these SVMs are processed by wavelet Inverse transform and reconstruction of hydrological time series, the establishment of wavelet analysis based on support vector machine hydrological trend analysis model to Sanmenxia Hydrological Station natural monthly runoff sequence as an example to verify the application. The results show that the proposed model is superior to the former in both prediction accuracy and length of time, compared with the traditional prediction models such as SVM and neural network.