پایش سلامت پل ها با استفاده از داده‌های موجود بر مبنای یادگیری عمیق

نوع مقاله : علمی - پژوهشی

نویسندگان

1 کارشناسی ارشد، دانشکده مهندسی عمران، دانشگاه علم و صنعت ایران، تهران، ایران

2 استاد، دانشگاه علم و صنعت، تهران، ایران

3 استادیار، گروه عمران، واحد رودهن، دانشگاه آزاد اسلامی، رودهن، ایران

چکیده

بروز آسیب‌های ناشی از بارگذاری‌های ناگهانی یا آسیب‌هایی در طول عمر سازه ممکن است باعث ایجاد تغییرات نامطلوبی در عملکرد آن ‌شود لذا در سال‌های اخیر نیاز به روش‌هایی جهت شناسایی آسیب درآن‌ها بیش از پیش احساس می‌شود. ازاین‌رو دست‌یابی به روش‌هایی جهت شناسایی رخداد و یافتن محل آسیب یکی از موضوعاتی است که همواره در مهندسی عمران، مکانیک و هوافضا مطرح بوده است. ایده اصلی در اکثر این روش‌ها استفاده از پاسخ‌هایی است که سازه بر اساس پیش‌فرض‌های موجود به عوامل بیرونی می‌دهد، در اغلب این مطالعات سازه به‌عنوان‌ یک سیستم ‌دینامیکی با جرم، سختی و میرایی یکتا در نظر گرفته می‌شود. و در صورت آسیب ‌این پارامترها با تغییر مواجه می‌شوند و این تغییرات خود را در پاسخ‌های سازه نشان می‌دهد. این پاسخ‌ها به‌صورت سیگنال‌های زمانی، حاوی خصوصیات دینامیکی سازه است که با تحلیل سیگنال استخراج می شوند و در تشخیص آسیب از آن‌ها بهره گرفته می شود. پیشرفت روزافزون علم یادگیری ماشین نه تنها شرایط را برای پیشرفت علوم مهندسی مهیا کرده است بلکه این پیشرفت به گونه‌ای بوده است که امروزه این کامپیوترها هستند که اغلب محاسبات مهم را بر عهده خواهند داشت، یکی از روش‌هایی که امروزه بسیار موردتوجه قرارگرفته است استفاده از شبکه‌های عصبی کانولوشن به‌منظور تشخیص خصوصیات ذکر شده در سیگنال‌های سازه است.
در این مقاله سعی شده است تا شبکه‌ی کانولوشن‌ یک‌بعدی جهت تشخیص حالت سالم از ناسالم طراحی شود، علاوه بر آن اثر تابع نرمال‌ساز در فاز پیش‌پردازش داده و اثر دو تابع بهینه‌ساز گرادیان کاهشی و آدام در فرآیند آموزش شبکه نیز موردبررسی قرار گرفته‌است. درنهایت می‌توان بیان کرد که نتایج حاکی از عملکرد رضایت بخش، دقیق و سریع روش شبکه عصبی عمیق کانولوشن یک‌بعدی در تشخیص حالت سالم از ناسالم است.

کلیدواژه‌ها

موضوعات


عنوان مقاله [English]

Health Monitoring of Bridges by Using the Available Data Based on Deep Learning

نویسندگان [English]

  • MohammadSadegh Torabi 1
  • Gholamreza Ghodrati Amiri 2
  • Ehsan Darvishan 3
1 M.Sc., Faculty of Civil Engineering, Iran University of Science and Technology, Tehran, Iran
2 Professor, Department of Civil Engineering, Iran University of Science & Technology, Tehran, Iran
3 Assistant Professor, Department of Civil Engineering, Roudehen Branch, Islamic Azad University, Roudehen, Iran
چکیده [English]

Damage caused by sudden and specific loads such as earthquakes causes undesirable changes in structural performance. Therefore, in recent years need for methods to identify damage in structures feel more than ever. Therefore, finding ways to identify damages and their location is one of the issues that has always been discussed in civil, mechanical, and aerospace engineering. The main idea in most of these methods is to use the responses that the structure gives to external factors based on the assumptions available, in most of these studies structures are considered as a dynamic system with unique mass, stiffness, and damping which in the case of structural damage, these parameters change and reflect changes in the structural responses. This response is often in the form of time signals containing structural properties that can be extracted and used to detect potential damage by examining them. The increasing advancement of machine learning science has not only provided the conditions for the improvement of engineering sciences but it has been such that today it is the computers that often do the most important calculations, One of the most widely used methods today is deep convolutional neural networks to detect the properties mentioned in structural signals.
This thesis attempts to design a one-dimensional convoluted network to detect a healthy state from unhealthy in which we investigate the effect of the normalizer function on the pre-processing phase is attempted and the effect of Stochastic decreasing gradient and Adam (two optimization functions) on the network learning process is also investigated. To evaluate the capability of the proposed method, data from the Yong cable bridge in China were used. Finally, it can be concluded that the results show satisfactory, accurate, and fast one-dimensional deep neural network (convolution) performance in the diagnosis of an unhealthy state.

کلیدواژه‌ها [English]

  • Structural health monitoring
  • damage detection
  • deep neural network
  • convolution neural network
  • ADAM optimization algorithm
[1]        J. Peng, S. Zhang, D. Peng, and K. Liang, “Application of machine learning method in bridge health monitoring,” 2017 2nd Int. Conf. Reliab. Syst. Eng. ICRSE 2017, no. Icrse, 2017, doi: 10.1109/ICRSE.2017.8030793.
[2]        K. H. Padil, N. Bakhary, and H. Hao, “The use of a non-probabilistic artificial neural network to consider uncertainties in vibration-based-damage detection,” Mech. Syst. Signal Process., vol. 83, pp. 194–209, Jan. 2017, doi: 10.1016/j.ymssp.2016.06.007.
[3]        P. da S. L. Alexandrino, G. F. Gomes, and S. S. Cunha, “A robust optimization for damage detection using multiobjective genetic algorithm, neural network and fuzzy decision making,” Inverse Probl. Sci. Eng., vol. 28, no. 1, pp. 21–46, Jan. 2020, doi: 10.1080/17415977.2019.1583225.
[4]        H. S. Kim, C. Jin, M. H. Kim, and K. Kim, “Damage detection of bottom-set gillnet using Artificial Neural Network,” Ocean Eng., vol. 208, p. 107423, Jul. 2020, doi: 10.1016/J.OCEANENG.2020.107423.
[5]        D. E. Sidarta, J. O’Sullivan, and H. J. Lim, “Damage detection of offshore platform mooring line using artificial neural network,” in Proceedings of the International Conference on Offshore Mechanics and Arctic Engineering - OMAE, Sep. 2018, vol. 1, doi: 10.1115/OMAE2018-77084.
[6]        A. Santos, E. Figueiredo, M. F. M. Silva, C. S. Sales, and J. C. W. A. Costa, “Machine learning algorithms for damage detection: Kernel-based approaches,” Elsevier, 2016.
[7]        C. Modarres, N. Astorga, E. L. Droguett, and V. Meruane, “Convolutional neural networks for automated damage recognition and damage type identification,” Struct. Control Heal. Monit., vol. 25, no. 10, pp. 1–17, 2018, doi: 10.1002/stc.2230.
[8]        M. Azimi, A. D. Eslamlou, and G. Pekcan, “Data-driven structural health monitoring and damage detection through deep learning: State-ofthe- art review,” Sensors (Switzerland), vol. 20, no. 10, 2020, doi: 10.3390/s20102778.
[9]        Y. Z. Lin, Z. H. Nie, and H. W. Ma, “Structural Damage Detection with Automatic Feature-Extraction through Deep Learning,” Comput. Civ. Infrastruct. Eng., vol. 32, no. 12, pp. 1025–1046, 2017, doi: 10.1111/mice.12313.
[10]      O. Abdeljaber, O. Avci, M. S. Kiranyaz, B. Boashash, H. Sodano, and D. J. Inman, “1-D CNNs for structural damage detection: Verification on a structural health monitoring benchmark data,” Neurocomputing, vol. 275, pp. 1308–1317, 2018, doi: 10.1016/j.neucom.2017.09.069.
[11]      Y. J. Cha, W. Choi, and O. Büyüköztürk, “Deep Learning-Based Crack Damage Detection Using Convolutional Neural Networks,” Comput. Civ. Infrastruct. Eng., vol. 32, no. 5, pp. 361–378, 2017, doi: 10.1111/mice.12263.
[12]      Y. Yu, C. Wang, X. Gu, and J. Li, “A novel deep learning-based method for damage identification of smart building structures,” Struct. Heal. Monit., vol. 18, no. 1, pp. 143–163, 2019, doi: 10.1177/1475921718804132.
[13]      Y. Bao, Z. Tang, H. Li, and Y. Zhang, “Computer vision and deep learning–based data anomaly detection method for structural health monitoring,” Struct. Heal. Monit., vol. 18, no. 2, pp. 401–421, 2019, doi: 10.1177/1475921718757405.
[14]      H. Liu and Y. Zhang, “Image-driven structural steel damage condition assessment method using deep learning algorithm,” Meas. J. Int. Meas. Confed., vol. 133, pp. 168–181, 2019, doi: 10.1016/j.measurement.2018.09.081.
[15]      T. Guo, L. Wu, C. Wang, and Z. Xu, “Damage detection in a novel deep-learning framework: a robust method for feature extraction,” Struct. Heal. Monit., no. 28, 2019, doi: 10.1177/1475921719846051.
[16]      H. Liu and Y. Zhang, “Deep learning-based brace damage detection for concentrically braced frame structures under seismic loadings,” Adv. Struct. Eng., 2019, doi: 10.1177/1369433219859389.
[17]      C. Modarres, N. Astorga, E. L. Droguett, and V. Meruane, “Convolutional neural networks for automated damage recognition and damage type identification,” Struct. Control Heal. Monit., vol. 25, no. 10, 2018, doi: 10.1002/stc.2230.
[18]      H. Li, J. Zhang, W. Li, and H. Jiang, “Study on the technology of replacement of closure segment for strengthening existing prestressed concrete cable stayed bridges,” Tumu Gongcheng Xuebao/China Civ. Eng. J., vol. 44, no. 7, pp. 83–89, 2011.
[19]      M. R. Kaloop and J. W. Hu, “Stayed-Cable Bridge Damage Detection and Localization Based on Accelerometer Health Monitoring Measurements,” Shock Vib., 2015, doi: 10.1155/2015/102680.
[20]      S. Arangio and F. Bontempi, “Structural health monitoring of a cable-stayed bridge with Bayesian neural networks,” Struct. Infrastruct. Eng., vol. 11, no. 4, pp. 575–587, 2015, doi: 10.1080/15732479.2014.951867.
[21]      J. DeLaughter, B. Meltz, S. Smith, J. Yun, and M. Murat, “Use of z-score to rescale amplitudes,” Lead. Edge, vol. 24, no. 7, pp. 698–701, 2005, doi: 10.1190/1.1993260.
[22]      Y. LeCun and others, “Generalization and network design strategies,” in Connectionism in perspective, 1989, pp. 143–155.
[23]      Y. LeCun et al., “Backpropagation Applied to Handwritten Zip Code Recognition,” Neural Comput., vol. 1, no. 4, pp. 541–551, 1989, doi: 10.1162/neco.1989.1.4.541.
[24]      E. R. Ziegel, “The Elements of Statistical Learning,” Technometrics, vol. 45, no. 3, pp. 267–268, 2003, doi: 10.1198/tech.2003.s770.
[25]      Y. T. Zhou and R. Chellappa, “Computation of optical flow using a neural network,” 1988, pp. 71–78, doi: 10.1109/icnn.1988.23914.
[26]      M. Heusel et al., “Fast and Accurate CNN Learning on ImageNet,” Iccv, p. 2015, 2015.
[27]      L. Bull, K. Worden, G. Manson, and N. Dervilis, “Active learning for semi-supervised structural health monitoring,” J. Sound Vib., vol. 437, pp. 373–388, 2018, doi: 10.1016/j.jsv.2018.08.040.
[28]      Y. Liu and Q. Liu, “Convolutional neural networks with large-margin softmax loss function for cognitive load recognition,” in Chinese Control Conference, CCC, 2017, pp. 4045–4049, doi: 10.23919/ChiCC.2017.8027991.
[29]      Z. Zhang and M. R. Sabuncu, “Generalized cross entropy loss for training deep neural networks with noisy labels,” in Advances in Neural Information Processing Systems, 2018, vol. 2018-Decem, pp. 8778–8788.
[30]      D. P. Kingma and J. Ba, “Adam: A Method for Stochastic Optimization,” pp. 1–15, 2014, [Online]. Available: http://arxiv.org/abs/1412.6980.