مطالعه موردی سیگنال‌های لرزه‌ای ایستگاه قیر قبل از رخداد زلزله

نوع مقاله : Articles

نویسنده

دانشگاه آزاد اسلامی، واحد تهران مرکزی، گروه برق و الکترونیک، تهران، ایران

چکیده

در این مقاله با استفاده از نتایج دو سری آزمون، مدل کردن آشوبناکی سیگنال بر اساس نویزها و تشخیص ناهنجاری‌های قبل از زلزله مورد مطالعه قرار گرفته است. ابتدا تعدادی ویژگی مانند ویژگی‌های فرکانسی، آماری و آشوب از سیگنال لرزه استخراج شد، سپس ماتریس ویژگی حاصل توسط شبکه پرسپترون چند لایه مورد بررسی قرار گرفت. شبکه پرسپترون توانایی تشخیص آشوبناکی بر اساس نویزها در فاصله زمانی 5 دقیقه قبل از وقوع زلزله با دقت قابل قبول 1404/81 درصد برای زلزله‌های 5-7 ریشتری ثبت شده سال‌های 2004-2010 برای 21 ایستگاه ایران را دارا می‌باشد. سیگنال‌های ثبت شده لرزه‌نگار ایستگاه قیر نیز با این روش مورد بررسی قرار گرفت، که شبکه آشوبناکی قبل از وقوع زلزله را با دقت قابل قبول 8696/60 درصد برای آن ایستگاه تشخیص داد. در مقاله حاضر تلاش می‌شود تا ویژگی‌های غیر وابسته به زمین‌شناسی از سیگنال‌های لرزه‌نگاشت یک ایستگاه و 21 ایستگاه استخراج شود، سپس نتایج شبکه پرسپترون در تشخیص ناهنجاری و آشوبناکی سیگنال بر اساس نویزها در فاصله زمانی 5 دقیقه قبل از زلزله برای هر دو حالت مورد مقایسه قرار گیرد. نتایج نشان می‌دهد که استفاده از تعداد بیشتر داده‌های ثبت شده هر ایستگاه و افزودن ویژگی‌های غیر زمین‌شناسی آنها به ویژگی‌های زمین‌شناختی هر منطقه می‌تواند در نتیجه شبکه پرسپترون برای مدل کردن آشوبناکی سیگنال بر اساس نویزها تأثیرگذار ‌باشد.

کلیدواژه‌ها


عنوان مقاله [English]

Case Study of Seismic Signals for Ghir Station before the Earthquake

نویسنده [English]

  • Leila Dehbozorgi
Department of Electrical & Electronic Engineering, Central Tehran Branch, Islamic Azad University, Tehran, Iran
چکیده [English]

In this paper, using two experimental results, chaotic signal modeling based on noise and detection of anomalies before an earthquake has been studied. First, a number of features such as frequency, statistical, entropy and FFT of the seismic signal were extracted, then the property matrix was investigated by multi-layered perceptron network (MLP). MLP can detect anomalies based on noise at a time interval of 5 minutes before an earthquake with an acceptable accuracy of 81.1404% for the magnitude 5-7 on the Richter scale from 21 earthquake recording stations in Iran (between 2004 and 2010). The recorded GHIR station signals were also investigated by this method, so MLP network determined the chaotic before an earthquake with an acceptable accuracy of 60.8696% for GHIR station. In this paper, it is tried to extract non-geologic properties from seismic signals of GHIR station and 21 stations. Then, the results of the MLP network are compared in terms of chaotic detection based on noise five minutes before an earthquake for each case. The results demonstrate that by using more recorded data of each station, MLP network is more capable for diagnosis of the chaotic based on noise before an earthquake.
The actual recorded seismic data is much better than simulated signals to detect seismic signal abnormalities before an earthquake and simulate the signal, in order to investigate the recorded signals of seismicity. In the recent study, recorded seismic signals were investigated.
The purpose of this study is to determine the most desirable characteristic matrix for detecting anomalies within five minutes before the earthquake.
Chaotic signal modeling based on noise and detection of anomalies before an earthquake using MLP network process contains four main stages: (1) Separate earthquake and slice signal, (2) The signals were processed using a high pass filter to reject the baseline drifts in signals, (3) Different types of features were extracted from the filtered signals and the feature vector was constructed, (4) The feature vector were fed to a neural network.
Our database contains 760 records for 21 stations and 148 records for GHIR station of IRAN earthquakes, from the International Institute of Earthquake Engineering and Seismology. The sampling frequency was 50 Hz.  . Fourth Order High Pass Butterworth Filter was used to annihilate low frequency (fc=0.04), and then the signal was normalized. The features were extracted from the filtered signals. Then the classifier was trained and tested. 70% of the data were randomly selected for training purposes and the rest 30% were used for testing for each experiment.
MLP networks were successfully trained using MATLAB (for both condition) and the results were presented for both conditions. The networks had one output in which every value at output uniquely representing each categories (1: earthquake, 0: no earthquake).
In this study, by using the results of two series of experiments, chaotic signal modeling based on noise and detection of anomalies before an earthquake has been investigated and the features matrix has been constructed for the following two conditions:

By using seismic signals in different geographic regions of the country (IRAN).
By using seismograph signals at a specific station (GHIR station).

The results show that MLP network produced better accuracy for 21 stations compared with GHIR station. Because the number of data was different in two ways or the features are non-geological. To achieve better results, the following suggestions are presented: noise reduction and use of more frequency features. We are now going to collect more and more earthquake data from all around the world, adding more chaos dependent parameters to the feature vector, using committee machines to increase the accuracy of classification and increasing prediction length from five minutes before to ten and more as well.

کلیدواژه‌ها [English]

  • Earthquake
  • Multilayer Perceptron Network
  • Wavelet transform
  • Chaotic Modeling Based on Noise
  • Lyapunov Exponent
  1. Zare, M. (2008) Evaluation of seismic attributes, Bam earthquake. The First Earthquake Precursor Workshop (in Persian).
  2. Çelık, E., Atalay, M., and Bayer, H. (2014) Earthquake prediction using seismic bumps with artificial neural networks and support vector machines. 22nd Signal Processing and Communications Applications Conference (SIU), 730-733.
  3. Alariï‌, A., Alariï‌, N.S.N., and Al-Humidan, S. (2011) Earthquakes magnitude predication using artiï‌cial neural network in northern Red Sea area. Journal of King Saud University, Science, 24, 301-313.
  4. Topcu, I.B. and Sarıdemir, M. (2008) Prediction of compressive strength of concrete containing ï‌‚y ash using artiï‌cial neural networks and fuzzy logic. Computational Materials Science, 305-311.
  5. Pazoki, A.R., Farokhi, F., and Pazoki, Z. (2014) Classification of rice grain varieties using two artificial neural networks (MLP and NEURO-FUZZY). The Journal of Animal & Plant Sciences, 24, 336-343.
  6. Nasrollah Nejad, A., Alamehzadeh, M., and Javan Doloyee, G. (2016) Estimation of the maximum peak ground acceleration by three artificial neural network models. 1-19 (in Persian).
  7. Dehbozorgi, L. (2014) Investigation of two groups of features for short-term Iran earthquake prediction using seismic signals with neuro-fuzzy classifier and multilayer perceptron neural network. The First National Conference on Metaheuristic Algorithms and its Applications in Science and Engineering (in Persian).
  8. Dehbozorgi, L. and Pazoki, Z. (2016) Study of the seismic signals before earthquakes larger than 5 on the Richter scale with neuro-fuzzy classifier and multilayer perceptron network. The 3rd International Congress on Technology Communication and Knowledge (in Persian).
  9. Podder, P., Hasan, M., Islam, R., and Sayeed, M. (2014) Design and implementation of Butterworth, ChebyshevI and elliptic filter for speech signal analysis. International Journal of Computer Applications. 98, 12-18.
  10. Noferesti, M. (2014) Statistics in business and economics. 23 (in Persian).
  11. Majumdar, K. and Myers M.H. (2006) Amplitude suppression and chaos control in epileptic EEG signals. Computational and Mathematical Methods in Medicine, 7, 53-66.
  12. Arbabi, E. () Investigation on relation between effective features used for different classifiers in BCI. M.S. thesis, Sharif Univ., Tehran (In Persian).
  13. Mohamad pour, M. and Farokhi, F. (2009) An advanced method for speech recognition. World Academy of Science, Engineering and Technology, 995-1000.
  14. Myint, S. W., Zhu, T., and Zheng, B. (2015) A novel image classification algorithm using over complete wavelet transforms. IEEE Geoscience and Remote Sensing, 12, 6.
  15. Scognamiglio, L. and Hutchings, L. (2009) A test of a physically-based strong ground motion prediction methodology with the 26 September 1997, Mw = 6.0 Colï‌orito (Umbria–Marche sequence), Italy earthquake. Tectonophysics, 145-158.
  16. Singh, N. and Khan, R. A. (2015) Speaker Recognition and Fast Fourier Transform. International Journal of Advanced Research in Computer Science and Software Engineering, 5(7), 530-534.
  17. Mack, C. A. (2015) More systematic errors in the measurement of power spectral density. J. Micro/Nanolith. MEMS MOEMS, 14(3), 033502-033508.
  18. Xue, Y., Ludovice, P. J., and Grover, M. A. (2011) Dynamic coarse graining in complex system simulation. American Control Conference on O'Farrell Street, 5031-5036.
  19. Abroudi, A., Farokhi, F., and Zahedi, F. (2012) Instance selection for training of intelligent networks based on fast condensed nearest neighbor rule. IEEE Conference, 1-4.
  20. Mathlab, [Online]. Available: https://www.mathworks.com
  21. International Institute of Earthquake Engineering and Seismology. [Online]. Available: http://www.iiees.ac.ir.