改进小波阈值降噪算法在水电机组信号处理中的应用
发布时间：2018-12-10 11:30
【摘要】：小波阈值降噪是近年来水电机组状态监测信号去噪常用的算法,经典的阈值函数为硬、软阈值两种函数。在硬阈值方法中,处理过的小波系数在阈值处是间断的,得到的估计信号在重构时易产生局部附加振荡;而经软阈值函数降噪的信号虽然整体连续性好,但与原始信号之间总是存在着恒定的偏差,影响重构精度。因此在软、硬阈值函数基础上提出了一种改进阈值函数的小波降噪算法,通过Matlab仿真和实际采集数据的实验结果表明:该方法克服了传统软、硬阈值函数算法的缺点,不仅连续性好,而且提高了信噪比。通过对比降噪后对振动信号特征能量的保留程度,说明改进阈值函数较传统经典去噪方法更为优越,是一种有效的降噪方法。
[Abstract]:Wavelet threshold de-noising is a commonly used algorithm for de-noising of state monitoring signals of hydropower units in recent years. The classical threshold functions are hard and soft threshold functions. In the hard threshold method, the processed wavelet coefficients are discontinuous at the threshold, and the estimated signals are prone to produce local additional oscillations during reconstruction. Although the signal denoised by soft threshold function has good overall continuity, there is always a constant deviation between the signal and the original signal, which affects the reconstruction accuracy. Therefore, a wavelet denoising algorithm based on soft and hard threshold functions is proposed. The results of Matlab simulation and actual data acquisition show that the method overcomes the shortcomings of the traditional soft and hard threshold function algorithms. Not only the continuity is good, but also the signal-to-noise ratio is improved. By comparing the reserve degree of characteristic energy of vibration signal after noise reduction, it is proved that the improved threshold function is more superior than the traditional classical denoising method, and it is an effective denoising method.
【作者单位】： 武汉大学动力与机械学院水电站过渡过程教育部重点实验室;
【基金】：国家自然科学基金项目(51379160)
【分类号】：TV734
[Abstract]:Wavelet threshold de-noising is a commonly used algorithm for de-noising of state monitoring signals of hydropower units in recent years. The classical threshold functions are hard and soft threshold functions. In the hard threshold method, the processed wavelet coefficients are discontinuous at the threshold, and the estimated signals are prone to produce local additional oscillations during reconstruction. Although the signal denoised by soft threshold function has good overall continuity, there is always a constant deviation between the signal and the original signal, which affects the reconstruction accuracy. Therefore, a wavelet denoising algorithm based on soft and hard threshold functions is proposed. The results of Matlab simulation and actual data acquisition show that the method overcomes the shortcomings of the traditional soft and hard threshold function algorithms. Not only the continuity is good, but also the signal-to-noise ratio is improved. By comparing the reserve degree of characteristic energy of vibration signal after noise reduction, it is proved that the improved threshold function is more superior than the traditional classical denoising method, and it is an effective denoising method.
【作者单位】： 武汉大学动力与机械学院水电站过渡过程教育部重点实验室;
【基金】：国家自然科学基金项目(51379160)
【分类号】：TV734
【参考文献】
相关期刊论文 前4条
1 唐友怀;张海涛;罗珊;姜U，
本文编号：2370494
本文链接：https://www.wllwen.com/kejilunwen/dianlilw/2370494.html
教材专著
