Ensemble models with neural networks for system identification
DOI:
https://doi.org/10.17979/ja-cea.2025.46.12205Keywords:
Nonlinear system identification, Neural networks, Identification and modelling, Ensemble models, Model validationAbstract
This work focuses on the development of complex dynamic models that can be integrated into a model predictive controller (MPC). The aim is to model nonlinear and multivariable systems using NARX neural networks (Nonlinear Autoregressive Networks with Exogenous Inputs). To improve the accuracy and robustness of the model obtained by the network, an ensemble modelling strategy is proposed, combining multiple independently trained networks. Two experiments are conducted: the first uses synthetic data of different natures to analyze the impact of diversity in the ensemble; the second one uses real data from a chemical reactor to assess the applicability of the approach in real-world multivariable environments. In both cases, it is shown that ensemble methods improve performance compared to individual models.
References
Barbez, A., Khomh, F., Gu´eh´eneuc, Y.-G., 3 2020. A machine-learning based ensemble method for anti-patterns detection. Journal of Systems and Software 161, 110486. DOI: 10.1016/j.jss.2019.110486
Barkhordari, M. S., Armaghani, D. J., Asteris, P. G., 2023. Structural damage identification using ensemble deep convolutional neural network models. Computer Modeling in Engineering & Sciences 134, 835–855. DOI: 10.32604/cmes.2022.020840
Cheng, A., Low, Y. M., 10 2023. Improved generalization of narx neural networks for enhanced metamodeling of nonlinear dynamic systems under stochastic excitations. Mechanical Systems and Signal Processing 200. DOI: 10.1016/j.ymssp.2023.110543
Fonollosa, J., 2015. Gas sensor array under dynamic gas mixtures. UCI Machine Learning Repository, DOI: https://doi.org/10.24432/C5WP4C.
Ganaie, M., Hu, M., Malik, A., Tanveer, M., Suganthan, P., 10 2022. Ensemble deep learning: A review. Engineering Applications of Artificial Intelligence 115, 105151. DOI: 10.1016/j.engappai.2022.105151
Harris, C. (Ed.), 1994. Advances in Intelligent Control. CRC Press.
Ibrahem, I. M., Akhrif, O., Moustapha, H., Staniszewski, M., 9 2021. Nonlinear generalized predictive controller based on ensemble of narx models for industrial gas turbine engine. Energy 230. DOI: 10.1016/j.energy.2021.120700
Lee, U., Kang, N., 7 2023. Adaptive neural network ensemble using prediction frequency. Journal of Computational Design and Engineering 10, 1547–1560. DOI: 10.1093/jcde/qwad071
Mohammed, A., Kora, R., 2 2023. A comprehensive review on ensemble deep learning: Opportunities and challenges. Leido el 08/10/2024. DOI: 10.1016/j.jksuci.2023.01.014
Nanni, L., Loreggia, A., Brahnam, S., 10 2023. Comparison of different methods for building ensembles of convolutional neural networks. Electronics 12, 4428. DOI: 10.3390/electronics12214428
Nugroho, H. A., Astuty, E. Y., Subiantoro, A., Kusumoputro, B., 2023. Ensemble deep learning narx for estimating time series of earthquake occurrence. In: 2023 3rd International Conference on Robotics, Automation and Artificial Intelligence, RAAI 2023. Institute of Electrical and Electronics Engineers Inc., pp. 301–305. DOI: 10.1109/RAAI59955.2023.10601278
Obregon, J., Jung, J. Y., 1 2022. Explanation of ensemble models. Elsevier Inc., pp. 51–72. DOI: 10.1016/B978-0-323-85648-5.00011-6
Roveda, L., Maskani, J., Franceschi, P., Abdi, A., Braghin, F., Tosatti, L. M., Pedrocchi, N., 11 2020. Model-based reinforcement learning variable impedance control for human-robot collaboration. Journal of Intelligent and Robotic Systems: Theory and Applications 100, 417–433. DOI: 10.1007/s10846-020-01183-3
Sasidaran, S., Raja, H. V., 2022. Recent trends in model predictive control. Tech. rep. URL: https://www.researchgate.net/publication/369201653
Schwenzer, M., Ay, M., Bergs, T., Abel, D., 11 2021. Review on model predictive control: an engineering perspective. DOI: 10.1007/s00170-021-07682-3
Wu, Z., Tran, A., Ren, Y. M., Barnes, C. S., Chen, S., Christofides, P. D., 5 2019. Model predictive control of phthalic anhydride synthesis in a fixedbed catalytic reactor via machine learning modeling. Chemical Engineering Research and Design 145, 173–183. DOI: 10.1016/j.cherd.2019.02.016
Downloads
Published
Issue
Section
License
Copyright (c) 2025 Ángel de la Peña, Eloy Irigoyen, Mikel Larrea
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
