Smart Structures and Systems

  • 제32권4호
  • /
  • Pages.267-279
  • /
  • 2023
  • /
  • 1738-1584(pISSN)
  • /
  • 1738-1991(eISSN)
테크노프레스 (Techno-Press)
권호 표지
DOI QR코드

DOI QR Code

Towards high-accuracy data modelling, uncertainty quantification and correlation analysis for SHM measurements during typhoon events using an improved most likely heteroscedastic Gaussian process

  • Qi-Ang Wang (State Key Laboratory of Intelligent Construction and Healthy Operation & Maintenance of Deep Underground Engineering & School of Mechanics and Civil Engineering, China University of Mining and Technology) ;
  • Hao-Bo Wang (State Key Laboratory of Intelligent Construction and Healthy Operation & Maintenance of Deep Underground Engineering & School of Mechanics and Civil Engineering, China University of Mining and Technology) ;
  • Zhan-Guo Ma (State Key Laboratory of Intelligent Construction and Healthy Operation & Maintenance of Deep Underground Engineering & School of Mechanics and Civil Engineering, China University of Mining and Technology) ;
  • Yi-Qing Ni (National Rail Transit Electrification and Automation Engineering Technology Research Center (Hong Kong Branch) and Department of Civil and Environmental Engineering, The Hong Kong Polytechnic University) ;
  • Zhi-Jun Liu (State Key Laboratory of Intelligent Construction and Healthy Operation & Maintenance of Deep Underground Engineering & School of Mechanics and Civil Engineering, China University of Mining and Technology) ;
  • Jian Jiang (State Key Laboratory of Intelligent Construction and Healthy Operation & Maintenance of Deep Underground Engineering & School of Mechanics and Civil Engineering, China University of Mining and Technology) ;
  • Rui Sun (State Key Laboratory of Intelligent Construction and Healthy Operation & Maintenance of Deep Underground Engineering & School of Mechanics and Civil Engineering, China University of Mining and Technology) ;
  • Hao-Wei Zhu (State Key Laboratory of Intelligent Construction and Healthy Operation & Maintenance of Deep Underground Engineering & School of Mechanics and Civil Engineering, China University of Mining and Technology)
  • 투고 : 2022.12.01
  • 심사 : 2023.10.12
  • 발행 : 2023.10.25
⟨ 이전 논문 다음 논문 ⟩

초록

Data modelling and interpretation for structural health monitoring (SHM) field data are critical for evaluating structural performance and quantifying the vulnerability of infrastructure systems. In order to improve the data modelling accuracy, and extend the application range from data regression analysis to out-of-sample forecasting analysis, an improved most likely heteroscedastic Gaussian process (iMLHGP) methodology is proposed in this study by the incorporation of the outof-sample forecasting algorithm. The proposed iMLHGP method overcomes this limitation of constant variance of Gaussian process (GP), and can be used for estimating non-stationary typhoon-induced response statistics with high volatility. The first attempt at performing data regression and forecasting analysis on structural responses using the proposed iMLHGP method has been presented by applying it to real-world filed SHM data from an instrumented cable-stay bridge during typhoon events. Uncertainty quantification and correlation analysis were also carried out to investigate the influence of typhoons on bridge strain data. Results show that the iMLHGP method has high accuracy in both regression and out-of-sample forecasting. The iMLHGP framework takes both data heteroscedasticity and accurate analytical processing of noise variance (replace with a point estimation on the most likely value) into account to avoid the intensive computational effort. According to uncertainty quantification and correlation analysis results, the uncertainties of strain measurements are affected by both traffic and wind speed. The overall change of bridge strain is affected by temperature, and the local fluctuation is greatly affected by wind speed in typhoon conditions.

키워드

과제정보

The study was supported by the National Key Research and Development Program of China under Award Number 2019YFE0118500, China Postdoctoral Science Foundation (2019M652006) and the National Natural Science Foundation of China (NSFC) under Award Numbers (51708545 and 52078478). The authors wish to express their gratitude to the staff and students in the Structural Engineering Laboratory for their extensive assistance. The data used to support the findings of this study are available from the corresponding author upon request.

참고문헌

  1. Agou, V.D., Pavlides, A. and Hristopulos, D.T. (2022), "Spatial modeling of precipitation based on data-driven warping of Gaussian processes", Entropy-Switz, 24(3), p. 321. https://doi.org/10.3390/e24030321
  2. Baesens, B., Viaene, S., Van den Poel, D., Vanthienen, J. and Dedene, G. (2002), "Bayesian neural network learning for repeat purchase modelling in direct marketing", Eur. J. Oper. Res., 138(1), 191-211. https://doi.org/10.1016/S0377-2217(01)00129-1
  3. Binois, M. and Gramacy, R.B. (2021), "hetGP: Heteroskedastic Gaussian Process Modeling and Sequential Design in R", J. Stat. Softw., 98(13), 1-44. https://doi.org/10.18637/jss.v098.i13
  4. Bludszuweit, H., Dominguez-Navarro, J.A. and Llombart, A. (2008), "Statistical analysis of wind power forecast error", IEEE Transact. Power Syst., 23(3), 983-991. https://doi.org/10.1109/TPWRS.2008.922526
  5. Braunfelds, J., Senkans, U., Skels, P., Janeliukstis, R., Porins, J., Spolitis, S. and Bobrovs, V. (2022), "Road Pavement Structural Health Monitoring by Embedded Fiber-Bragg-Grating-Based Optical Sensors", Sensors, 22(12). https://doi.org/10.3390/s22124581
  6. Brownjohn, J.M.W. (2007), "Structural health monitoring of civil infrastructure", Philosoph. Transact. Royal Soc. A: Mathe. Phys. Eng. Sci., 365(1851), 589-622. https://doi.org/10.1098/rsta.2006.1925
  7. Crawford, L., Wood, K.C., Zhou, X. and Mukherjee, S. (2019), "Bayesian Approximate Kernel Regression With Variable Selection", J. Am. Stat. Assoc., 113(524), 1710-1721. https://doi.org/10.1080/01621459.2017.1361830
  8. Glisic, B. (2022), "Concise historic overview of strain sensors used in the monitoring of civil structures: The first one hundred years", Sensors, 22(6), p. 2397. https://doi.org/10.3390/s22062397
  9. Goldberg, P.W. (1998), "Regression with input-dependent noise: A Gaussian process treatment", Adv. Neural Inform. Process. Syst., 10, 493-499.
  10. Herbko, M., Lopato, P., Psuj, G. and Rajagopal, P. (2022), "Application of Selected Fractal Geometry Resonators in Microstrip Strain Sensors", IEEE Sens. J., 22(13), 12656-12663. https://doi.org/10.1109/JSEN.2022.3177932
  11. Iba, Y. and Akaho, S. (2010), "Gaussian process regression with measurement error", IEICE Transact. Inform. Syst., 93(10), 589-622. https://doi.org/10.1587/transinf.E93.D.2680
  12. Ju, M., Park, C. and Kim, G. (2011), "Structural Health Monitoring (SHM) for a cable stayed bridge under typhoon", KSCE J. Civil Eng., 19(4), 1058-1068. https://doi.org/10.1007/s12205-015-0039-3
  13. Kersting, K. (2007), "Most Likely Heteroscedastic Gaussian Process Regression", Proceedings of the 24th International Conference on Machine Learning, pp. 393-400.
  14. Khan, M.S., Caprani, C., Ghosh, S. and Ghosh, J. (2021), "Value of strain-based structural health monitoring as decision support for heavy load access to bridges", Struct. Infrastruct. Eng., 18(4), 521-536. https://doi.org/10.1080/15732479.2021.1890140
  15. Ko, J.M. and Ni, Y.Q. (2005), "Technology developments in structural health monitoring of large-scale bridges", J. Eng. Struct., 27(12), 1715-1725. https://doi.org/10.1016/j.engstruct.2005.02.021
  16. Kopsaftopoulos. F.P. and Fassois, S.D. (2011), "Scalar and vector time series methods for vibration based damage diagnosis in a scale aircraft skeleton structure", J. Theor. Appl. Mech., 49(3), 727-756. https://doi.org/10.1088/1742-6596/305/1/012056
  17. Lee, W.F., Cheng, T.T., Huang, C.K., Yen, C.I. and Mei, H.T. (2014), "Performance of a highway bridge under extreme natural hazards: Case study on bridge performance during the 2009 Typhoon Morakot", J. Perform. Constr. Facil., 28(1), 49-60. https://doi.org/10.1061/(ASCE)CF.1943-5509.0000418
  18. Locke, W., Sybrandt, J., Redmond, L., Safro, I. and Atamturktur, S. (2020), "Using drive-by health monitoring to detect bridge damage considering environmental and operational effects", J. Sound Vib., 468. https://doi.org/10.1016/j.jsv.2019.115088
  19. Majkovic, D., O'Kiely, P., Kramberger, B., Vracko, M., Turk, J., Pazek, K. and Rozman, C. (2016), "Comparison of using regression modeling and an artificial neural network for herbage dry matter yield forecasting", J. Chemometr., 30(4), 203-209. https://doi.org/10.1002/cem.2770
  20. Mohammadi, M., Al-Fuqaha, A., Sorour, S and Guizani, M. (2007), "Deep learning for IoT big data and streaming analytics: A survey", IEEE Commun. Surv. Tutorials, 20(4), 2923-2960. https://doi.org/10.1109/COMST.2018.2844341
  21. Mousavi, M. and Gandomi, A.H. (2021), "Structural health monitoring under environmental and operational variations using MCD prediction error", J. Sound Vib., 512. https://doi.org/10.1016/j.jsv.2021.116370
  22. Munoz-Gonzalez, L., Lazaro-Gredilla, M. and Figueiras-Vidal, A.R. (2016), "Laplace approximation for divisive gaussian processes for nonstationary regression", IEEE T. Pattern Anal., 38(3), 618-624. https://doi.org/10.1109/TPAMI.2015.2452914
  23. Ni, Y.Q., Xia, Y., Liao, W.Y. and Ko, J.M. (2009), "Technology innovation in developing the structural health monitoring system for Guangzhou new TV tower", Struct. Control Health Monitor., 16(1), 73-98. https://doi.org/10.1002/stc.303
  24. Papadimitriou, C. (2004), "Optimal sensor placement methodology for parametric identification of structural systems", J. Sound Vib., 278(4-5), 923-947. https://doi.org/10.1016/j.jsv.2003.10.063
  25. Parida, S.S., Nikellis, A., Sett, K. and Singla, P. (2020), "Model-data fusion for seismic performance evaluation of an instrumented highway bridge", Earthq. Eng. Struct. Dyn., 49(14), 1559-1578. https://doi.org/10.1002/eqe.3317
  26. Partal, T. (2017), "Wavelet regression and wavelet neural network models for forecasting monthly streamflow", J. Water Clim. Change, 8(1), 48-61. https://doi.org/10.2166/wcc.2016.091
  27. Porter, K.A., Beck, J.L. and Shaikhutdinov, R.V. (2002), "Sensitivity of building loss estimates to major uncertain variables", Earthq. Spectra, 18(4), 719-743. https://doi.org/10.1193/1.1516201
  28. Rasmussen, C.E. and Nickisch, H. (2010), "Gaussian Processes for Machine Learning (GPML) Toolbox", J. Mach. Learn Res., 11, 3011-3015. https://doi.org/10.1115/1.4002474
  29. Sierra-Garcia, J.E. and Santos, M. (2021), "Improving wind turbine pitch control by effective wind neuro-estimators", IEEE Access., 9, 10513-10425. https://doi.org/10.1109/ACCESS.2021.3051063
  30. Skolidis, G. and Sanguinetti, G. (2011), "Bayesian Multitask Classification with Gaussian Process Priors", IEEE T. Neural Networ., 22(12), 2011-2021. https://doi.org/10.1109/TNN.2011.2168568
  31. Subasi, A. (2013), "Classification of EMG signals using PSO optimized SVM for diagnosis of neuromuscular disorders", Comput. Biol. Med., 43(5), 576-586. https://doi.org/10.1016/j.compbiomed.2013.01.020
  32. Talaei, S. and Ma, H.W. (2007), "Vibration-based Structural Damage Detection Using Twin Gaussian Process (TGP)", Structures, 16, 10-19. https://doi.org/10.1016/j.istruc.2018.08.006
  33. Theiler, M., Frangi, A. and Steiger, R. (2014), "Strain-based calculation model for centrically and eccentrically loaded timber columns", Eng. Struct., 56, 1103-1116. https://doi.org/10.1016/j.engstruct.2013.06.032
  34. Urban, S., Ludersdorfer, M. and van der Smagt, P. (2015), "Sensor Calibration and Hysteresis Compensation with Heteroscedastic Gaussian Processes", IEEE Sens. J., 15(11), 6498-6506. https://doi.org/10.1109/JSEN.2015.2455814
  35. Von Krannichfeldt, L., Wang, Y and Hug, G. (2007), "Online Ensemble Learning for Load Forecasting", IEEE T. Power Syst., 36(1), 545-648. https://doi.org/10.1109/TPWRS.2020.3036230
  36. Wan, H.P. and Ni, Y.Q. (2019a), "Binary Segmentation for Structural Condition Classification Using Structural Health Monitoring Data", J. Aerospace Eng., 32(1). https://doi.org/10.1061/(ASCE)AS.1943-5525.0000956
  37. Wan, H.P. and Ni, Y.Q. (2019b), "Bayesian modeling approach for forecast of structural stress response using structural health monitoring data", J. Struct. Eng., 144(9). https://doi.org/10.1061/(ASCE)ST.1943-541X.0002085
  38. Wan, H.P. and Ni, Y.Q. (2019c), "Bayesian multi-task learning methodology for reconstruction of structural health monitoring data", Struct. Health Monit., 18(4), 1282-1309. https://doi.org/10.1177/1475921718794953
  39. Wan, H.P., Dong, G.S., Luo, Y.Z. and Ni, Y.Q. (2022), "An improved complex multi-task Bayesian compressive sensing approach for compression and reconstruction of SHM data", Mech. Syst. Signal Pr., 167. https://doi.org/10.1016/J.YMSSP.2021.108531
  40. Wang, Q.A., Wu, Z. and Liu, S. (2018a), "Multivariate probabilistic seismic demand model for the bridge multidimensional fragility analysis", KSCE J. Civil Eng., 22(9), 3443-3451. https://doi.org/10.1007/s12205-018-1792-4
  41. Wang, Q.A., Wu, Z. and Liu, S. (2018b), "Multivariate probabilistic seismic demand model for the bridge multidimensional fragility analysis", KSCE J. Civil Eng., 22(9), 3443-3451. https://doi.org/10.1007/s12205-018-1792-4
  42. Wang, Q.A., Wang, C.B., Ma, Z.G., Chen, W., Ni, Y.Q., Wang, C.F., Yan, B.G. and Guan, P.X. (2022a), "Bayesian dynamic linear model framework for SHM data forecasting and missing data imputation during typhoon events", Struct. Health Monitor., 21(6), 2933-2950. https://doi.org/10.1177/14759217221079529
  43. Wang, Q.A., Zhang, C., Ma, Z.G., Jiao, G.Y., Jiang, X.W., Ni, Y.Q., Wang, Y.C., Du, Y.T., Qu, G.B. and Huang, J. (2022b), "Towards long-transmission-distance and semi-active wireless strain sensing enabled by dual-interrogation-mode RFID technology", Struct. Control Health., 29(11), e3069. https://doi.org/10.1002/stc.3069
  44. Wang, Q.A., Zhang, C., Ma, Z.G. and Ni, Y.Q. (2022c), "Modelling and forecasting of SHM strain measurement for a large-scale suspension bridge during typhoon events using variational heteroscedasic Gaussian process", Eng. Struct., 251, 113554. https://doi.org/10.1016/j.engstruct.2021.113554
  45. Wang, Q.A., Dai, Y., Ma, Z.G., Wang, J.F., Lin, J.F., Ni, Y.Q., Ren, W.X., Jiang, J., Yang, X. and Yan, J.R. (2023), "Towards high-precision data modeling of SHM measurements using an improved sparse Bayesian learning scheme with strong generalization ability", Struct. Health Monitor. https://doi.org/10.1177/14759217231170316
  46. Wright, J.H. (2008), "Bayesian Model Averaging and exchange rate forecasts", J. Econometr., 146(2), 329-341. https://doi.org/10.1016/j.jeconom.2008.08.012
  47. Yoon, S. and Kim, S. (2010), "k-Top Scoring Pair Algorithm for feature selection in SVM with applications to microarray data classification", Soft Comput., 14(2), 151-159. https://doi.org/10.1007/s00500-009-0437-x
  48. Zhang, Q.H. and Ni, Y.Q. (2020), "Improved most likely heteroscedastic Gaussian process regression via Bayesian residual moment estimator", IEEE Transact. Signal Process., 68, 3450-3460. https://doi.org/10.1109/TSP.2020.2997940