论文部分内容阅读
经验模态分解方法由于缺少数据预处理,分解结果具有冗余性、与自变量数量不能对应、模函数无法进行物理解释等问题,因此在经验模态分解前增加移动平均数据预处理过程,以减少原始数据量,保留趋势过程,并以某双曲拱坝位移分离为例,对比了经过数据预处理与未经数据预处理的经验模态分解结果。结果表明,预处理后位移序列趋势性不变,但数据量和分解模函数数量均有所减少;为了使分解得到的模函数与水位和温度的物理解释相吻合,选择合适的移动平均参数M,使模函数数量为2,避免了分解的冗余性。
Due to the lack of data preprocessing, the empirical mode decomposition method is redundant, the number of independent variables can not correspond to, and the modulo function can not be physically explained. Therefore, the moving average data preprocessing is added before empirical mode decomposition Reduce the amount of the original data, and keep the trend. Taking the displacement of a double-arch dam as an example, we compare the empirical mode decomposition results with and without data preprocessing. The results show that the trend of the displacement sequence does not change after preconditioning, but the amount of data and the number of decomposition modulus are all reduced. In order to make the modulus function of decomposition consistent with the physical interpretation of water level and temperature, select the appropriate moving average parameter M , The number of modulo function is 2, avoiding the redundancy of decomposition.