【摘要】 为提高大坝变形预测精度，针对大坝原始监测信号中的噪声，以及其非平稳性、非线性等特点，引入奇异谱分析(SSA)和局部均值分解(LMD)方法，提出SSA-LMD-GM模型。采用奇异谱分析(SSA)对原始监测信号进行去噪处理，为充分提取大坝形变信息特征，利用局部均值分解(LMD)对去噪后的监测信号进行分解。针对乘积函数(PF)分量的特征采用合适的模型预测分析，剩下余项则采用GM(1,1)模型。利用实际工程案例进行检验，结果表明，相较于其他模型，SSA-LMD-GM模型预测精度和拟合精度更加优秀，能较好地预测大坝变形趋势，具有一定的应用价值。更多还原

【Abstract】 In order to improve the prediction accuracy of dam deformation, singular spectrum analysis(SSA) and local mean decomposition(LMD) methods are introduced to the prediction model of dam deformation. Singular spectrum analysis(SSA) is used to denoise the original monitoring signals, and local mean decomposition(LMD) is used to decompose the de-noised monitoring signals so as to fully extract the deformation characteristics of the dam. The product function(PF) component is predicted by appropriate model, and the GM(1,1) model is used for the rest. The results show that compared with other models, SSA-LMD-GM model has better prediction accuracy and fitting accuracy, and can better predict the deformation trend of dam, which has certain application value.更多还原