Wind and Structures

Techno-Press (테크노프레스)
권호 표지
DOI QR코드

DOI QR Code

Enhancement of durability of tall buildings by using deep-learning-based predictions of wind-induced pressure

  • K.R. Sri Preethaa (Department of Robot and Smart System Engineering, Kyungpook National University) ;
  • N. Yuvaraj (Department of Robot and Smart System Engineering, Kyungpook National University) ;
  • Gitanjali Wadhwa (Department of Computer Science and Engineering, KPR Institute of Engineering and Technology) ;
  • Sujeen Song (Earth Turbine) ;
  • Se-Woon Choi (Department of Architectural Engineering, Daegu Catholic University) ;
  • Bubryur Kim (Department of Robot and Smart System Engineering, Kyungpook National University)
  • Received : 2022.11.23
  • Accepted : 2023.03.31
  • Published : 2023.04.25
⟨ Previous Next ⟩

Abstract

The emergence of high-rise buildings has necessitated frequent structural health monitoring and maintenance for safety reasons. Wind causes damage and structural changes on tall structures; thus, safe structures should be designed. The pressure developed on tall buildings has been utilized in previous research studies to assess the impacts of wind on structures. The wind tunnel test is a primary research method commonly used to quantify the aerodynamic characteristics of high-rise buildings. Wind pressure is measured by placing pressure sensor taps at different locations on tall buildings, and the collected data are used for analysis. However, sensors may malfunction and produce erroneous data; these data losses make it difficult to analyze aerodynamic properties. Therefore, it is essential to generate missing data relative to the original data obtained from neighboring pressure sensor taps at various intervals. This study proposes a deep learning-based, deep convolutional generative adversarial network (DCGAN) to restore missing data associated with faulty pressure sensors installed on high-rise buildings. The performance of the proposed DCGAN is validated by using a standard imputation model known as the generative adversarial imputation network (GAIN). The average mean-square error (AMSE) and average R-squared (ARSE) are used as performance metrics. The calculated ARSE values by DCGAN on the building model's front, backside, left, and right sides are 0.970, 0.972, 0.984 and 0.978, respectively. The AMSE produced by DCGAN on four sides of the building model is 0.008, 0.010, 0.015 and 0.014. The average standard deviation of the actual measures of the pressure sensors on four sides of the model were 0.1738, 0.1758, 0.2234 and 0.2278. The average standard deviation of the pressure values generated by the proposed DCGAN imputation model was closer to that of the measured actual with values of 0.1736,0.1746,0.2191, and 0.2239 on four sides, respectively. In comparison, the standard deviation of the values predicted by GAIN are 0.1726,0.1735,0.2161, and 0.2209, which is far from actual values. The results demonstrate that DCGAN model fits better for data imputation than the GAIN model with improved accuracy and fewer error rates. Additionally, the DCGAN is utilized to estimate the wind pressure in regions of buildings where no pressure sensor taps are available; the model yielded greater prediction accuracy than GAIN.

Keywords

Acknowledgement

This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (No. 2022R1A2C1009781). This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (No. 2022R1C1C1005409).

References

  1. Afrifa-Yamoah, E., Mueller, U.A., Taylor, S.M. and Fisher, A.J. (2022), "Missing data imputation of high-resolution temporal climate time series data", Meteorol. Appl., 27(1), 1873. https://doi.org/10.1002/met.1873. 
  2. Alkesaiberi, A. Harrou, F. and Sun, Y. (2022), "Efficient wind power prediction using machine learning methods: a comparative study", Energies, 15, 2327. https://doi.org/10.3390/en15072327. 
  3. Kim, B., Tse, K.T., Yoshida, A., Chen, Z., Van Phuc, P. and Park, H.S. (2019), "Investigation of flow visualization around linked tall buildings with circular sections", Build. Environ., 153, 60-76. https://doi.org/10.1016/j.buildenv.2019.02.021. 
  4. Kim, B., Tse, K.T., Yoshida, A., Tamura, Y., Chen, Z., Van Phuc, P. and Park, H.S. (2019), "Statistical analysis of wind-induced pressure fields and PIV measurements on two buildings", J. Wind Eng. Ind. Aerod., 188, 161-174, https://doi.org/10.1016/j.jweia.2019.01.016. 
  5. Kim, B., Tse, K.T., Chen, Z. and Park, H.S. (2020), "Multi-objective optimization of a structural link for a linked tall building system", J. Build. Eng., 31, 101382. https://doi.org/10.1016/j.jobe.2020.101382. 
  6. Bubryur Kim, N. Yuvaraj, K.T. Tse, Dong-Eun Lee, Gang Hu (2021), "Pressure pattern recognition in buildings using an unsupervised machine-learning algorithm", J. Wind Eng. Ind. Aerodyn., 214, 104629, https://doi.org/10.1016/j.jweia.2021.104629. 
  7. Dong, M., Wu, H., Hu, H., Azzam, R., Zhang, L., Zheng, Z. and Gong, X. (2020), "Deformation prediction of unstable slopes based on real-time monitoring and deepar model", Sensors, 21, 14. https://doi.org/10.3390/s21010014. 
  8. Bre, F., Gimenez, J.M. and Fachinotti, V.D. (2018), "Prediction of wind pressure coefficients on building surfaces using artificial neural networks", Energy Build., 158, 1429-1441. https://doi.org/10.1016/j.enbuild.2017.11.045. 
  9. Fan, H. Zhang, X. Mei, S. Chen, K. and Chen, X. (2020), "M2GSNet: multi-modal multi-task graph spatiotemporal network for ultra-short-term wind farm cluster power prediction", Appl. Sci., 10, 7915. https://doi.org/10.3390/app10217915. 
  10. Fan, H., Zhang, X. and Mei, S. (2021), "Wind power time series missing data imputation based on generative adversarial network", IEEE 4th International Electrical and Energy Conference, 1-6. https://doi.org/10.1109/CIEEC50170.2021.9510923. 
  11. Hu, G., Liu, L., Tao, D., Song, J., Tse, K.T. and Kwok, K.C. (2020), "Deep learning- based investigation of wind pressures on tall building under interference effects", J. Wind Eng. Ind. Aerod., 201, 104138, https://doi.org/10.1016/j.jweia.2020.104138. 
  12. Goodfellow, I., Pouget-Abadie, J., Mirza, M., Xu, B., Warde-Farley, D., Ozair, S. and Bengio, Y. (2020), "Generative adversarial networks", Communications of the ACM, 63(11), 139-144. https://doi.org/10.48550/arXiv.1406.2661. 
  13. Huang, F., Zhang, J. and Zhou, C. (2020), "A deep learning algorithm using a fully connected sparse autoencoder neural network for landslide susceptibility prediction", Landslides, 17, 217-229. https://doi.org/10.1007/s10346-019-01274-9. 
  14. Kim, B., Yuvaraj, N., Sri Preethaa, K.R., Hu, G. and Lee, D.E. (2021), "Wind-Induced pressure prediction on tall buildings using generative adversarial imputation network", Sensors, 21, 2515. https://doi.org/10.3390/s21072515. 
  15. Lee, S.J. and Yoon, H.K. (2021), "Discontinuity predictions of porosity and hydraulic conductivity based on electrical resistivity in slopes through deep learning algorithms", Sensors, 21, 1412. https://doi.org/10.3390/s21041412. 
  16. Mir, A.A., Kearfott, K.J., Celebi, F.V. and Rafique, M. (2022), "Imputation by feature importance (IBFI): A methodology to envelop machine learning method for imputing missing patterns in time series data", PLoS ONE, 17(1), e0262131. https://doi.org/10.1371/journal.pone.0262131. 
  17. Tse, K.T., Hitchcock, P.A. and Kwok, K.C.S. (2009), "Mode shape linearization for HFBB analysis of wind-excited complex tall buildings", Eng. Struct., 31(3), 675-685, https://doi.org/10.1016/j.engstruct.2008.11.012. 
  18. Niu, Y., Fritzen, C.P., Jung, H., Buethe, I., Ni, Y.Q. and Wang, Y.W. (2015), "Online simultaneous reconstruction of wind load and structural responses-theory and application to canton tower", Comput. Aid. Civil Infrastr. Eng., 30(8), 666-681, https://doi.org/10.1111/mice.12134. 
  19. Oh, B.K., Glisic, B., Kim, Y. and Park, H.S. (2019), "Convolutional neural network-based wind-induced response estimation model for tall buildings", Computer-Aided Civil and Infrastr. Eng., 34(10), 843-858. https://doi.org/10.1111/mice.12476. 
  20. Salinas, D., Flunkert, V., Gasthaus, J. and Januschowski, T. (2020), "DeepAR: probabilistic forecasting with autoregressive recurrent networks", Int. J. Forecast., 36, 1181-1191. https://doi.org/10.1016/j.ijforecast.2019.07.001. 
  21. Thakker, D., Mishra, B.K., Abdullatif, A., Mazumdar, S. and Simpson, S. (2020), "Explainable artificial intelligence for developing smart cities solutions", Smart Cities, 3, 1353-1382. https://doi.org/10.3390/smartcities3040065. 
  22. Thrun, M.C., Ultsch, A. and Breuer, L. (2021), "Explainable AI framework for multivariate hydro chemical time series", Mach. Learn. Knowl. Extr., 3, 170-204. https://doi.org/10.3390/make3010009. 
  23. Thrun, M.C. and Ultsch, A. (2021), "Using projection-based clustering to find distance- and density-based clusters in high-dimensional data", J. Classification, 38, 280-312. https://doi.org/10.1007/s00357-020-09373-2. 
  24. Park, H., Yuvaraj, N. and Kim, B. (2019), "Statistical analysis of aerodynamic characteristics on buildings", J. Korean Assoc. Spatial Struct., 19(4), 14-18, https://doi.org/10.1007/s00357-020-09373-2. 
  25. Wesonga, R. (2015), "On multivariate imputation and forecasting of decadal wind speed missing data", SpringerPlus, 4, 12. https://doi.org/10.1186/s40064-014-0774-9.