混合小波去噪在瞬变电磁资料处理中的应用

Application of mixed wavelet denoising method in TEM data processing

为提高小波阈值去噪在瞬变电磁资料处理中的效果,将多种改进的小波阈值函数和传统小波阈值函数对各实测TEM数据进行滤波处理,再按一定的标准筛选出各点最优的阈值处理函数作为该点最终的去噪函数。结果表明:对比混合小波和单一小波函数去噪的效果发现,混合小波去噪在瞬变电磁资料处理中能更有效压制噪声,突出信号异常,去噪效果优于单一的阈值去噪方法。通过改进该混合小波函数去噪的代码还可以筛选出适合各点去噪的最优小波基,使混合小波去噪更加完善。

This paper is an effort to improve the effect of wavelet threshold denoising on transient electromagnetic data processing. The effort is directed at applying a variety of improved wavelet threshold functions and the traditional wavelet threshold functions for filtering all the measured TEM data; and ac- cording to a certain criterion, selecting the best optimal threshold function of each point as the final de- noising function. The comparison between the denoising effect of mixed wavelet and that of single improved wavelet reveals that, when applied in transient electromagnetic data processing, the mixed wavelet method may afford a better denoising effect than the single threshold denoising one, thanks to its more efective noise suppression and better abnormal signal clarification. An improvement in the mixed wavelet denoising function allows the selection of the optimal wavelet base suitable for each point, thus contributing to a better performance of the mixed wavelet denoising method.