IncepTCN: a New Deep Temporal Convolutional Network Combined With Dictionary Learning for Strong Cultural Noise Elimination of Controlled-Source Electromagnetic Data
| Autor: | Guang Li, Shouli Wu, Hongzhu Cai, Zhushi He, Xiaoqiong Liu, Cong Zhou, Jingtian Tang |
|---|---|
| Rok vydání: | 2023 |
| Předmět: | |
| Zdroj: | GEOPHYSICS. :1-68 |
| ISSN: | 1942-2156 0016-8033 |
| Popis: | When the controlled-source electromagnetic (CSEM) data are contaminated by intense cultural noise and the signal-to-noise ratio (SNR) is lower than 0 dB, the existing denoising methods can hardly achieve good results. To overcome the problem, a new strong-noise elimination method called IncepTCN-SISC is proposed based on deep learning and dictionary learning. First, a novel deep neural network (DNN) model called IncepTCN is created based on the Inception block and temporal convolutional network (TCN). Then, we employ IncepTCN to recognize the strong-noise segments from the observed signal, which will be discarded later. Finally, a dictionary learning method based on shift-invariant convolutional coding is used to denoise the remaining weak-noise segments. Through a series of simulated and field data experiments, we find that the new proposed IncepTCN network has obvious advantages over the comparison methods in accuracy and efficiency. The average recognition accuracy of IncepTCN is 96.5%, which is 25.5%, 3.2%, 1.1%, and 2.0% higher than that of the fuzzy C-means clustering (FCM), convolutional neural network (CNN), residual network (ResNet), and the non-improved TCN, respectively. In addition, the test results of unfamiliar data show that the generalization ability of IncepTCN is significantly better than the CNN, ResNet, and non-improved TCN. The proposed IncepTCN-SISC method can improve the SNR of CSEM data from -5.0 dB to 3.1 dB or from 5.0 dB to 31.9 dB and solve the denoising problem of noisy data below 0 dB to a certain extent. After IncepTCN-SISC processing, the initially distorted apparent resistivity curves become smooth, and the result is better than dictionary learning. The proposed method is intelligent without any manual intervention and is suitable for batch processing of CSEM data. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|