Manullang, Dillon Amangda Barry (2020) KOREKSI DAN FILTERING DATA REKAMAN SEISMIK DENGAN MENGGUNAKAN TEKNIK EMPIRICAL MODE DECOMPOSITION. Undergraduate thesis, Institut Teknologi Sepuluh Nopember.
Preview |
Text
03411640000049-Undergraduate_Thesis.pdf Download (3MB) | Preview |
Abstract
Metode seimsik merupakan salah satu metode geofisika yang digunakan untuk mendeleniasi bawah permukaan. Dalam aplikasi di lapangan, sinyal hasil akuisisi seismik tidak selalu terbebas dari noise. Penelitian ini bertujuan untuk membenamkan noise acak dan koheren serta menghilangkan trend pada data rekaman seismik dengan menggunakan teknik Empirical Mode Decomposition(EMD). Penelitian ini menggunakan 4 variasi data sintetis dan 2 data rekaman mikroseismik . Data atau sinyal didekomposisi menggunakan teknik EMD sehingga menghasilkan kumpulan subsinyal yang disebut Intrinsic Mode Function (IMF). IMF ini dianalisa keterkaitanya dengan noise dan trend dalam data secara kualitatif dan kuantitatif . Hasil analisa ini menghasilkan tahapan yang digunakan untuk denoising dan koreksi baseline berdasarkan dekomposisi EMD. Kinerja metode ini dievaluasi dengan menggunakan parameter SNR,MAE/MSE dan Cross Correlation. Hasil evaluasi kinerja pada data sintetis dan lapangan mengindikasikan bahwa denoising dan detrending berdasarkan dekomposisi EMD yang dilakukan cukup efektif namun pada data lapangan dipelukan analisis parameter secara kualitatif guna meningkatkan kualitas hasil filtering. Rata rata nilai Mean Absolute Error(MAE) koreksi baseline dengan menggunakan teknik EMD adalah sebesar 0.0044, rata-rata Mean Squared Error(MSE) sebesar 0.0449, Signal to Noise Ratio(SNR) sebesar 7.4902 dB dan Korelasi sebesar 0.8340 sedangkan nilai MAE pada denoising menggunakan EMD Adaptive interval tresholding adalah sebesar 5.3000e-05, MSE sebesar 0.0046, SNR sebesar 19.1444 dB dan Korelasi sebesar 0.9935.
==============================================================Seimsic method is one of the geophysical methods used to delineate subsurface. In field applications, seismic acquisition signals are not always noise free. This study aims to immerse random and coherent noise and eliminate trends in seismic recording data using the Empirical Mode Decomposition (EMD) technique. This research uses 4 variations of synthetic data and 2 microseismic data records. Data or signals are decomposed using the EMD technique to produce a collection of subscriptions named the Intrinsic Mode Function (IMF). The IMF is analyzed to find the relationship of noise and trends in data qualitatively and quantitatively. The results of this analysis produce stages used for denoising and baseline correction based on EMD decomposition. The performance of this method is evaluated using the SNR, MAE / MSE and Cross Correlation parameters. The results of performance evaluation on synthetic data and field data indicate that denoising and detrending based on EMD decomposition is quite effective but needs qualitative analysisis to enhance the filtering result . With the average MAE value on the baseline correction using EMD technique is 0.0044, the average MSE is 0.0449, the SNR is 7.4902 dB and the Correlation is 0.8340 while the MAE value at denoising using EMD Adaptive interval tresholding is at 5.3000e-05, MSE is at 7,4902 dB and Correlation is at 0.8340 while the MAE value at denoising using EMD Adaptive interval tresholding is at 5.3000e-05, MSE at 0.0046, SNR of 19.1444 dB and Correlation of 0.9935.
| Item Type: | Thesis (Undergraduate) |
|---|---|
| Uncontrolled Keywords: | Empirical Mode Decomposition, Intrinsic Mode Function, Noise,Trend |
| Subjects: | Q Science > QE Geology > QE538.5 Seismic tomography; Seismic waves. Elastic waves T Technology > T Technology (General) > T57.5 Data Processing |
| Divisions: | Faculty of Civil, Planning, and Geo Engineering (CIVPLAN) > Geophysics Engineering > 33201-(S1) Undergraduate Thesis |
| Depositing User: | Dillon Amangda Barry Manullang |
| Date Deposited: | 23 Aug 2020 15:07 |
| Last Modified: | 17 May 2023 02:38 |
| URI: | http://repository.its.ac.id/id/eprint/79376 |
Actions (login required)
 |
View Item |