Vol. 19, No. 4, pp. 366-374 (2023)
INVERSION METHOD OF UNCERTAIN PARAMETERS FOR TRUSS
STRUCTURES BASED ON GRAPH NEURAL NETWORKS
Zhang-Qi Wang 1, 2, 3, Zhe Zheng 1, Jun-Wei MengXiang 1 and Wen-Qiang Jiang 1, 2, 3, *
1 Department of Mechanical Engineering, North China Electric Power University, Baoding, China
2 Hebei Engineering Research Center for Advanced Manufacturing & Intelligent Operation and Maintenance of Electric Power Machinery,
North China Electric Power University, Baoding, China
3 Hebei Key Laboratory of Electric Machinery Health Maintenance & Failure Prevention, North China Electric Power University, Baoding 071003, China
*(Corresponding author: E-mail:This email address is being protected from spambots. You need JavaScript enabled to view it.)
Received: 7 April 2023; Revised: 6 June 2023; Accepted: 20 June 2023
DOI:10.18057/IJASC.2023.19.4.5
 View Article |
Export Citation: Plain Text | RIS | Endnote |
ABSTRACT
Uncertainty exists widely in practical engineering. It is an important challenge in engineering structural analysis. In truss structures, the uncertainties of axial stiffness of bolted joints will significantly affect the mechanical behavior of the structure as the axial load is dominated by the member internal forces. Structural response analysis based on determined structural parameters is a common forward problem that can be solved by modeling analysis methods. However, the uncertainties parameter of axial stiffness of bolted joint cannot be determined during the design and analysis of truss structure in the direct nonlinear analysis method. Structural parameter identification based on structural response is a typical inverse problem in engineering, which is difficult to solve using traditional analysis tools. In this paper, an inverse model based on Graph Neural Network (GNN) is proposed. The feature encoding method for transforming truss structures into graph representations of GNN is defined. A parameterized acquisition method for large-scale datasets is presented, and an innovative inversion model based on GNN for the inversion of uncertain parameters of truss structures is proposed. The proposed method is shown to perform well with an inversion accuracy, and accurate results can be obtained with limited data sets. The inversion method has strong data mining capability and model interpretability, making it a promising direction for exploring engineering structural analysis.
KEYWORDS
Truss structures, Uncertain parameters, Inversion method, GNN
REFERENCES
[1] Gallet A., Rigby S., Tallman T. N., et.al., “Structural engineering from an inverse problems perspective”, Proceedings. Mathematical, Physical and Engineering Sciences, 2022, Vol. 478, No. 2257, pp. 1-37.
[2] Zhao Z. W., Liu H. Q., Liang B., et.al., “Influence of random geometrical imperfection on the stability of single-layer reticulated domes with semi-rigid connection”, Advanced Steel Construction, 2019, Vol. 15, No. 1, pp. 93-99.
[3] Srinivasan K., Selvakumar K., “Comparative study of pre-engineered building and space truss building with different span”, Indian Journal of Engineering and Materials Sciences, 2022, Vol. 29, No. 4, pp. 428-431.
[4] Fairclough H. E., Gilbert M., Pichugin A. V., et.al., “Theoretically optimal forms for very long-span bridges under gravity loading”, Proceedings. Mathematical, Physical and Engineering Sciences, 2018, Vol. 474, No. 2217, pp. 1-21.
[5] He G., Shao X. D., Chen Y. B., et.al., “Preliminary design of a steel-UHPFRC composite truss arch bridge and model tests of K-joints”, Journal of Bridge Engineering, 2022, Vol. 27, No. 10, 04022090.
[6] Grzywinski M., “Optimization of spatial truss towers based on Rao algorithms”, Structural Engineering and Mechanics, 2022, Vol. 81, No. 3, pp. 367-378.
[7] Jiang W. Q., Liu Y. P., Chan S. L., Wang Z. Q., “Direct analysis of an ultrahigh-voltage lattice transmission tower considering joint effects”, Journal of Structural Engineering, 2017, Vol. 143, No. 5, pp.04017009.1-04017009.14.
[8] Wang F. Y., Xu Y. L., Zhan S., “Concurrent multi-scale modeling of a transmission tower structure and its experimental verification”, Advanced Steel Construction, 2017, Vol. 13, No.3, pp. 258-272.
[9] Hussain A., Liu Y. P., Chan S. L., “Finite element modeling and design of single angle member under bi-axial bending”, Structures, 2018, Vol. 16, pp. 373-389.
[10] Chan S. L., Fong M., “Experimental and analytical investigations of steel and composite trusses”, Advanced Steel Construction, 2011, Vol. 7, No. 1, pp. 17-26.
[11] An L. Q., Wu J., Jiang W. Q., “Experimental and numerical study of the axial stiffness of bolted joints in steel lattice transmission tower legs”, Engineering Structures, 2019, Vol. 187, pp. 490-503.
[12] Li J. X., Cheng J. P., Zhang C., et.al., “Seismic response study of a steel lattice transmission tower considering the hysteresis characteristics of bolt joint slippage”, Engineering Structures, 2023, Vol. 281, 115754.1-115754.15.
[13] Habeck M., “Bayesian approach to inverse statistical mechanics”, Physical Review E, 2014, Vol. 89, No. 5, pp. 052113.1-052113.7.
[14] Adler J., Öktem O., “Solving ill-posed inverse problems using iterative deep neural networks”, Inverse Problems, 2017, Vol. 33, 124007.
[15] Kamyab S., Azimifar Z., Sabzi R., Fieguth P., “Deep learning methods for inverse problems”, PeerJ. Computer Science, 2022, Vol. 8, e951.
[16] Han X., Liu J., Chen J. L., “A manifold learning method for inverse problems with structural multi-source uncertainties”, Chinese Journal of Computational Mechanics, 2021, Vol. 38, No. 4, pp. 523-530.
[17] Keshavarzzadeh V., Kirby R.M., Narayan A., “Variational inference for nonlinear inverse problems via neural net kernels: Comparison to Bayesian neural networks, application to topology optimization”, Computer Methods in Applied Mechanics and Engineering, 2022, Vol. 400, pp.1-35.
[18] Jagtap A.D., Mao Z.P., Adams N., Karniadakis G.E., “Physics-informed neural networks for inverse problems in supersonic flows”, Journal of Computational Physics, 2022, Vol. 466, pp. 1-19.
[19] Raissi M., Perdikaris P., Karniadakis G.E., “Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations”, Journal of Computational Physics, 2019, Vol. 378, pp. 686-707.
[20] Wang C., Fan J. S., “A general intelligent computation framework for structural responses in civil engineering based on deep learning”, Journal of Building Structures, 2023, Vol. 44, No. 1, pp. 259-268.
[21] Battaglia P. W., Hamrick J. B., Bapst V., et.al., “Relational inductive biases, deep learning, and graph networks”,2018, ArXiv Preprint ArXiv: 1806.01261.
[22] Fout A., Byrd J., Shariat B., et.al., “Protein interface prediction using graph convolutional networks”, In Proc. of NIPS, 2017, Vol. 30, pp. 6530-6539.
[23] Hamaguchi T., Oiwa H., Shimbo M., et.al., “Knowledge transfer for out-of-knowledge-base entities: A graph neural network approach”, In Proc. of IJCAI, 2017, pp. 1802-1808.
[24] Battaglia P., Pascanu R., Lai M., et.al., “Interaction networks for learning about objects, relations and physics”, In Proc. of NIPS, 2016, Vol. 29, pp. 4502-4510.
[25] Sanchez A., Heess N., Springenberg J. T., et.al., “Graph networks as learnable physics engines for inference and control”, In Proc. of ICLR, 2018, Vol. 80, pp. 4467-4476.
[26] Zhao Z. W., Liang B., Liu H. Q., Li Y. J., “Simplified numerical model for high-strength bolted connections. Engineering Structures”, 2018, Vol. 164, pp. 119-127.
[27] Wang D., Zhang Z. S., “A four-parameter model for nonlinear stiffness of a bolted joint with non-Gaussian surfaces”, Acta Mechanica, 2020, Vol. 231, No. 5, pp. 1963-1976.
[28] Li C. F., Jiang Y. L., Qiao R. H., Miao X. Y., “Modeling and parameters identification of the connection interface of bolted joints based on an improved micro-slip model”, Mechanical Systems and Signal Processing, 2021, Vol. 153, pp. 1-16.
[29] Jiang W. Q., Wang Z. Q., McClure G., et.al., “Accurate modeling of joint effects in lattice transmission towers”, Engineer Structure, 2011, Vol. 33, No. 5, pp. 1817–27.