Identification of Impact Loads and Partial System Parameters Using 1D-CNN

Authors: Xuewen Yu, Danhui Dan

Abstract:

The identification of impact loads and some hard-to-obtain system parameters is crucial for analysis, validation, and evaluation activities in the engineering field. This paper proposes a method based on 1D-CNN to identify impact loads and partial system parameters from the measured responses. To this end, forward computations are conducted to provide datasets consisting of triples (parameter θ, input u, output y). Two neural networks are then trained: one to learn the mapping from output y to input u and another to learn the mapping from input and output (u, y) to parameter θ. Subsequently, by feeding the measured output response into the trained neural networks, the input impact load and system parameter can be calculated, respectively. The method is tested on two simulated examples and shows sound accuracy in estimating the impact load (waveform and location) and system parameter.

Keywords: Convolutional neural network, impact load identification, system parameter identification, inverse problem.

References:

[1] R. Liu, E. Dobriban, Z. Hou, and K. Qian, ”Dynamic load identification for mechanical systems: a review,” Arch. Comput. Methods Eng., 1–33, 2021.
[2] T.S. Jang, H. Baek, S.L. Han, and T. Kinoshita, ”Indirect measurement of the impulsive load to a nonlinear system from dynamic responses: inverse problem formulation,” Mech. Syst. Sig. Process., 24 (6), 1665–1681, 2010.
[3] H. Inoue, N. Ikeda, K. Kishimoto, T. Shibuya, and T. Koizumi, ”Inverse analysis of the magnitude and direction of impact force,” JSME Int. J. Ser. A, Mech. Mater. Eng., 38 (1), 84–91, 1994.
[4] G. Yan, H. Sun, and O. Büyüköztürk, ”Impact load identification for composite structures using Bayesian regularization and unscented Kalman filter,” Struct. Control Health Monit., 24, e1910, 2017.
[5] J. Jiang, H. Tang, M.S. Mohamed, S. Luo, and J. Chen, ”Augmented Tikhonov Regularization Method for Dynamic Load Identification,” Appl. Sci., 10, 6348, 2020.
[6] Y. Li, X. Wang, Y. Xia, and L. Sun, ”Sparse Bayesian technique for load identification and full response reconstruction,” J. Sound Vib., 553, 117669, 2023.
[7] Y. Liu, L. Wang, ”A two-step weighting regularization method for stochastic excitation identification under multi-source uncertainties based on response superposition-decomposition principle,” Mech. Syst. Signal Process., 182, 109565, 2023.
[8] B. Qiao, X. Zhang, C. Wang, H. Zhang, and X. Chen, ”Sparse regularization for force identification using dictionaries,” J. Sound Vib., 368, 71–86, 2016.
[9] J. Liu, X. Sun, X. Han, C. Jiang, and D. Yu, ”A novel computational inverse technique for load identification using the shape function method of moving least square fitting,” Comput. Struct., 144, 127–137, 2014.
[10] G. Yan and L. Zhou, ”Impact load identification of composite structure using genetic algorithms,” J. Sound Vib., 319, 869–884, 2009.
[11] R. Hashemi and M. Kargarnovin, ”Vibration base identification of impact force using genetic algorithm,” Int. J. Mech. Syst. Sci. Eng., 1 (4), 204–210, 2007.
[12] J. Zhou, L. Dong, W. Guan, J. Yan, ”Impact load identification of nonlinear structures using deep Recurrent Neural Network,” Mech. Syst. Signal Process., 133, 106292, 2019.
[13] J. Li, J. Yan, J. Zhu, and X. Qing, ”K-BP neural network-based strain field inversion and load identification for CFRP,” Measurement, 187, 110227, 2022.
[14] S.-M. Baek, J.-C. Park, and H.-J. Jung, ”Impact load identification method based on artificial neural network for submerged floating tunnel under collision,” Ocean Eng., 286, 115641, 2023.
[15] C. Guo, L. Jiang, F. Yang, Z. Yang, and X. Zhang, ”Impact load identification and localization method on thin-walled cylinders using machine learning,” Smart Mater. Struct., 32, 065018, 2023.
[16] D. P., Kingma, and J. Ba, ”Adam: A method for stochastic optimization,” arXiv preprint, arXiv:1412.6980, 2014.