提出了一种基于小波变换和支持向量机的大坝变形预测方法。通过小波变换把变形时间序列分解成具有不同频率特征的分量,根据各分量的特点构造不同的支持向量机模型进行预测,然后把各分量的预测结果进行重构,作为最终的变形预测结果。实例证明,该方法具有很高的预测精度和较强的泛化能力。
A novel model based on wavelet transform and support vector machine for dam deformation prediction is presented. Firstly, through the wavelet transform, deformation time series is decomposed into different frequency components. Then, according to the different characteristics of the decomposed components, different support vector machines are constructed to forecast the components. Finally, the predicted results of the components are reconstructed to be used as the final prediction result of deformation. The calculation result shows that this model has higher forecasting precision and greater generality ability.