论文部分内容阅读
针对现有变形预测方法对于大坝变形的预测效果不理想的问题,该文利用局部均值分解方法获取生产函数分量并进行支持向量回归建模,用此方法对大坝变形进行多尺度分析。通过局部均值分解对大坝变形序列进行分解得到其乘积函数分量,然后利用支持向量机回归进行外推预测,再把各乘积函数分量的预测结果进行叠加重构生成,进而获得大坝变形预测值。通过实例分析,比较GM(1,1)、支持向量机和该文方法3种模型在变形监测数据处理中的拟合和预测结果,表明该文方法充分发掘数据本身所蕴含的物理机制和物理规律,提高了大坝变形多尺度预测精度。
Aiming at the problem that the existing deformation forecasting method is not ideal for the prediction of dam deformation, this paper uses the local average decomposition method to obtain the production function components and support vector regression modeling, and uses this method to conduct multi-scale analysis of the dam deformation. Through the local average decomposition, the deformation sequence of the dam is decomposed to obtain the product function component, and then the support vector machine regression is used to extrapolate the prediction. The prediction results of the product function components are then superimposed and reconstructed to obtain the dam deformation prediction value . Through case analysis, the fitting and forecasting results of GM (1,1), support vector machine (SVM) and the proposed method are compared in deformation monitoring data processing. The results show that the method fully explores the physical mechanism and physics The law improves the multi-scale prediction accuracy of dam deformation.