تخمین مقادیرحداکثر شتاب جنبش نیرومند زمین توسط سه مدل شبکه های عصبی مصنوعی

نوع مقاله : Articles

نویسندگان

1 پژوهشگاه بین‌المللی زلزله‌شناسی و مهندسی زلزله، تهران

2 پژوهشکده زلزله شناسی، پژوهشگاه بین المللی زلزله شناسی و مهندسی زلزله، تهران

چکیده

در این تحقیق، برای برآورد بیشینه شتاب جنبش نیرومند زمین در یک منطقه، از سه نوع شبکه‌ی عصبی مصنوعی با الگوریتم‌های متفاوت استفاده شده است که عبارتند از: شبکه‌ی عصبی- فازی هم فعال، المان– جردن و پیشخور با الگوریتم پس انتشار خطا. بردارهای ورودی شبکه‌های عصبی، شامل چهار پارامتری هستند که تأثیرات مهمی در وقوع زمین‌لرزه در یک منطقه دارند. این پارامترها عبارتند از: بزرگی ممان زلزله، شعاع گسیختگی کانون زلزله، مکانیسم گسل و رده‌بندی ساختگاه. بردار خروجی نیز فقط یک مؤلفه دارد: حداکثر شتاب جنبش نیرومند زمین برای یک زمین‌لرزه رخ داده در یک منطقه که به‌عنوان خروجی هدف استفاده می‌شود. پس از انجام آزمایش‌های مختلف، از میان شبکه‌های عصبی طراحی‌شده، شبکه‌ی عصبی- فازی هم فعال (سی‌آنفیس) بالاترین ضریب همبستگی خروجی، برابر 82/0 و شبکه پیشخور با الگوریتم پس انتشار خطای عمومی، کمترین ضریب همبستگی 41/0 را نشان می‌دهد. همچنین شبکه سی‌آنفیس، کمترین میانگین مربعات خطای 075/0 و شبکه‌ی پیشخور، بیشترین میانگین مربعات خطای 125/0 را داشته است. در این تحقیق، شبکه عصبی – فازی هم فعال، بهترین شبکه‌ی عصبی است که می‌تواند حداکثر دامنه شتاب احتمالی بالاتر از g1 در یک منطقه را برآورد کند.

کلیدواژه‌ها


عنوان مقاله [English]

Estimating Values of the Maximum Peak Ground Acceleration of a Strong Motion by Three Models of Artificial Neural Networks

نویسندگان [English]

  • Ali Nasrollahnejad 1
  • Mosafa Allamehzadeh 2
  • Gholam Javan Doloei 2
1 International Institute of Earthquake Engineering and Seismology (IIEES), Tehran, Iran
2 International Institute of Earthquake Engineering and Seismology (IIEES), Tehran, Iran
چکیده [English]

Peak ground acceleration is one of the most important factors that needs to be investigated in order to predict the devastation potential resulting from earthquakes in reconstruction sites. Besides, the maximum level of shaking control is subjected criteria that can be worth considering. In this research, a training algorithm based on gradient descent and Levenberg-Marquart (Train LM) were developed and employed by using strong ground motion records. The Artificial Neural Networks (ANN) algorithm indicated that the fitting between the predicted PGA values by the networks and the observed PGA values were able to yield high correlation coefficients of 0.78 for PGA.
From a deterministic point of view, the determination of the strongest level of shaking that is expected at a site has long been a significant consideration in earthquake engineering. Besides, knowledge of the maximum physically possible ground motions allows a meaningful truncation of the distribution of ground motion residuals, and as a result, leads to falling of the values computed in probabilistic seismic hazard analysis (Strasser and Bommer, 2009).
The peak ground acceleration parameter is often estimated by the attenuation of relationships and by using regression analysis. PGA is one of the most important parameters, often analyzed in studies related to damages caused by earthquakes (Gullo and Ercelebi, 2007). It is mostly estimated by the attenuation of equations and is developed by a regression analysis of powerful motion data.
Kerh and Chaw (2002) used software calculation techniques to remove the lack of certainties in declining relations. They used the mixed gradient training algorithm of Fletcher-Reeves’ back propagation error (Fletcher and Reeves, 1964). They applied three neural network models with different inputs including epicentric distance, focal depth and magnitude of the earthquakes. These records were trained and then the output results were compared with available nonlinear regression analysis.
In this article, to estimate strong ground motion acceleration component in an area, four artificial neural networks with different algorithms were used, including General Regression Neural Network (GRNN), Nonlinear Auto Regression neural network (NARX), Feed-Forward Back-Propagation error (FFBP) and General Feed-Forward Neural Network (GFFNN). Input vectors of neural networks include four parameters, which have key effects in occurrence of an earthquake in an area. The parameters include magnitude of moment, rupture distance of earthquake center, mechanism of faults, and ranking of site. Output vector has only one component: maximum peak ground acceleration for an earthquake in an area is used as a target output.
After different tests, GRNN network has maximum output correlation coefficient (0.87) and General Feed-forward Back-Propagation error neural network (FFBP) has the least (0.41). Besides, GRNN network had the least mean square error (0/014), and Back-Propagation network had 0.125. In this research, GRNN neural network is the best neural network, which can estimate possible peak acceleration more than 1g in an area.
Artificial neural networks are a set of non-linear optimizer methods which do not need certain mathematical models in order to solve problems. In regression analysis, PGA is calculated as a function of earthquake magnitude, distance from the source of the earthquake to the site under study, local condition of the site and other characteristics that are linked to the earthquake source such as slippery length and reverse, normal or wave propagation. In non-linear regression methods, non-linear relations that exist between input and output parameters are expressed as estimations, through statistical calculations within a specified relationship (Douglas, 2003).

کلیدواژه‌ها [English]

  • Maximum Peak Ground Acceleration
  • Moment Magnitude
  • General Feed-Forward Neural Network
  • Ranking of Sites
  • Elman-Jordan Neural Network
  1. Gullo, H. and Ercelebi, E. (2007) A neural network approach for attenuation relationships. Engineering Geology, 93, 65-81.
  2. Kreh, T. and Chaw, D. (2002) Neural network approach and micro tremor measurements in estimating peak ground acceleration due to strong motion. Advances in Engineering Software, 33, 733-742.
  3. Strasser, F.O., Bommer, J.J., and Abrahamson, N.A. (2008) Truncation of the distribution of ground motion residuals. Journal of Seismology, 12(1), 79-105.
  4. Strasser, F.O. and Bommer, J.J. (2009) Large-amplitude ground-motion recordings and their interpretations. Soil Dynamics and Earthquake Engineering, 29, 1305-1329.
  5. Moradzadeh, A. and Bakhshi, E. (2006) Simulation of photoelectric log in oil-bearing formation using artificial neural network. Journal of Earth and Space Physic, 32, 1-20 (in Persian).
  6. Cartalopus, S.V. (2003) Fuzzy Logic and Neural Networks, Applications and Concepts.
  7. Gunaydin, K. and Gunaydin, A. (2008) Peak ground acceleration prediction by artificial neural network for northwestern Turkey. Mathematical Problems in Engineering.
  8. Kia, S.M. (2011) Neural Networks in Matlab. Qian Academic Publishing, 408 pages.
  9. McGarr, A., Green, R.W.E., and Spottiswood, S.M. (1981) Strong ground motion mine tremors: some implications for near-source ground motion parameters. Bulletin of the Seismological Society of America, 71(1), 295-319.
  10. Shakal, A.F., Haddadi, H.R., Graizer, V., Lin, K., and Huang, M. (2006) Some key features of the strong motion data from the M 6.0 Parkfield, California, earthquake of 28 September 2004. Bulletin of the Seismological Society of America, 96(4B), S90-S118.
  11. Shing, J. and Jang, R. (1993) ANFIS: adaptive network based fuzzy inference systems. IEEE Transactions on Systems, 23(3).
  12. Werbos, P. J. (1974) Beyond Regression, New Tools for Prediction and Analyses in the Behavioral Sciences. Ph.D. Thesis, Harward University.
  13. Jang, J.S.R., Sun, C.T., and Mizutani, E. (1997) Neuro-Fuzzy and Soft Computing. NJ, Prentice-Hall.
  14. Haykin, S. (1999) Neural Networks, a Comprehensive Foundation. Prentice Hall, Englewood Cliffs, NJ, USA, 2nd Edition.
  15. Altun, S., Guktepe, A.B., Ansal, A.M., and Kguner, C.A. (2009) Simulation of torsional shear testresults with neuro-fuzzy control system. Soil Dynamic and Earthquake Engineering, 29, 253-260.
  16. Jang, J.S.R and Sun, C.T. (1995) Neuro-Fuzzy modeling and control. Proc. IEEE, 83(3), 378-405.
  17. Tsukamoto, Y. (1979) ‘An approach to reasoning method’. In: Advances in Fuzzy Set Theory and Application, Gupta, M., Ragade, R.K., Yager, R.R. (Eds.), 137-149.
  18. Rumelhart, D.E. and Mc.Clelland, J.L. (1986) Parallel Distributed Processing. Vol I, II, MIT, Cambridge.
  19. Krose, B. and Smagt, P.V. (1996) An Introduction to Neural Network. University of Amsterdam, 47-50.
  20. Jordan, M.I. (1986) Attractor Dynamics and parallelism in a connectionist sequential machine. Proceeding of the 8th Annual Conference of the Cognitive Science Society, 531-546.
  21. Elman, J.L. (1990) Finding structure in time. Cognitive Science, 14, 179-211.