基于小波包-神经网络混合算法的瞬变电磁信号降噪研究  

Denoising of transient electromagnetic data based on wavelet packet-neural network algorithm

作　　者：古瑶 解海军[1] 李璐 李刚 GU Yao;XIE Haijun;LI Lu;LI Gang(School of Geology and Environment,Xi'an University of Science and Technology,Xi'an 715400,China;Key Laboratory of Coal Resources Exploration and Comprehensive Utilization,Xi'an 710048,China)

机构地区：[1]西安科技大学地质与环境学院,陕西西安715400 [2]国土资源部煤炭资源勘查与综合利用重点实验室,陕西西安710048

出　　处：《煤炭工程》2023年第5期147-152,共6页Coal Engineering

基　　金：陕西省自然科学基础研究计划企业联合基金资助项目(2021JLM-11)。

摘　　要：针对瞬变电磁信号容易受到电磁干扰的影响,使得数据信噪比降低,衰减曲线失真,而单一的滤波方法存在易丢失地质信息、圆滑过度等缺点,因而较难获得高精度成像结果的问题。提出基于小波包变换-BP神经网络的混合算法,利用小波包变换对信号的能量特征提取能力、分解重构能力和BP神经网络的学习和反馈能力,对瞬变电磁信号进行滤波处理。利用傅里叶变换获得瞬变电磁信号的频域特征,将受干扰信号和未受干扰信号对比,得到两者差异;利用3层小波包分解获得第3层节点能量占比,提取重构信号特征,对瞬变电磁信号进行初步的分解重构;调用训练好的神经网络模型对重构信号进行特征训练,获得最终的瞬变电磁信号。经小保当矿业有限公司二号煤矿实测数据研究表明:混合算法相比常用的S-G滤波、均值滤波等,实用性更强、准确率更高,滤波时保留了真实的地质信息,增强了资料解释的精度,应用效果较好,为资料处理过程提供了强有力的技术支持。Transient electromagnetic signals are susceptible to electromagnetic interference,which reduces the signal-to-noise ratio of data and distorts the attenuation curve.However,single-filtering method has the disadvantages of geological information loss and excessive smoothness,so it is difficult to obtain high-precision imaging.Therefore,a hybrid algorithm based on wavelet packet transform and BP neural network is proposed,which uses wavelet packet transform's ability to extract energy features,decompose and reconstruct signals and BP neural network's learning and feedback ability to filter transient electromagnetic signals.Fourier transform is used to obtain the frequency domain characteristics of the transient electromagnetic signal,and the difference between the interfered and the undisturbed signals are obtained.Three-layer wavelet packet decomposition is used to obtain the energy ratio of the third-layer nodes,extract the characteristics of the reconstructed signal,and preliminarily decompose and reconstruct the transient electromagnetic signal.The trained neural network model is called to train the features of the reconstructed signal,and the final filtered transient electromagnetic signal is obtained.The theoretical and measured data research shows that the hybrid algorithm is more practical and accurate than the commonly used S-G filtering and mean filtering.It keeps the real geological information while filtering,enhances the accuracy of data interpretation,and has a good application effect,which provides a strong technical support for data processing.

关 键 词：瞬变电磁法 小波包变换 BP神经网络 混合算法 降噪 

分 类 号：TD676[矿业工程—矿山机电]