大坝水平位移监测数据的小波变换去噪处理

将实际观测到的一组大坝水平位移监测数据作为通常意义下的时序信号，并对其进行小波变换分析。根据噪声和有用信号的小波变换系数模极大值在不同分解尺度上的传播特点，提出了根据模极大值的传播规律来区别噪声和有用信号的方法，并给出了具体的算法和计算程序。对模拟数据和实测数据的处理结果表明，基于小波分解模极大值的去噪方法能够有效剔除土木工程变形监测数据中的噪声，识别被噪声湮没的有用信号。

WAVELET TRANSFORM ANALYSIS AND DENOISING OF MONITORING DATA OF HORIZONTAL DAM DISPLACEMENT

A set of monitoring data of horizontal dam displacement is treated and analyzed as a time series in a common sense and its wavelet transform is analyzed. Based on the propagation characteristics of wavelet transform coefficient modulus maxima across different scales, a new method as well as its algorithm and computational procedure are provided for distinguishing noises from valuable signals. The results obtained from both synthetic signals and real data indicate that the denoising method based on the modulus maxima of wavelet transform can effectively reject the noise in the data of deformation monitoring in civil engineering and identify weak signals corrupted seriously by noise.