MODELS OF DYNAMIC OBJECTS WITH SIGNIFICANT NONLINEARITY BASED ON TIME-DELAY NEURAL NETWORKS
DOI:
https://doi.org/10.24025/2306-4412.3.2023.288284Keywords:
identification, nonlinear objects, substantial nonlinearities, dynamic neural networks, simulation modelingAbstract
The paper is devoted to modeling nonlinear objects using dynamic neural networks. The aim of the work is to improve the accuracy of modeling dynamic objects with significant nonlinearities using neural network models and to determine the scope of effective application of these models. This aim is achieved using time-delay neural networks. To assess the applicability of the proposed models, the study considers simulation objects with two types of nonlinearities: smooth and piecewise linear (saturation). The investigation of suggested models accuracy in nonlinear dynamic object modeling involves experiments: assessing the models' scalability with different input signals; evaluating their extrapolation capabilities. The findings from these experiments are compared with the modeling results using the compensatory method of deterministic identification based on functional series. The outcomes of the simulation experiments reveal that the suggested models lack invariance concerning the input signal. Nevertheless, when trained on a comprehensive dataset of representative input signals, neural network models can effectively capture the characteristics of nonlinear dynamic objects. The extrapolation abilities of the proposed models tend to degrade as the input signal amplitudes exceed the range covered by the used training set. The scientific novelty is the establishment of a clear relationship between the types of input signals, their amplitudes, and the accuracy of the proposed models. The practical significance of investigation delineates the areas in which time-delay neural networks can be used to address the real-world challenges associated with significantly non-linear objects; demonstrates increasing the accuracy of identifying nonlinear objects compared to functional series models.
References
Agresti, A. (2017). Foundations of linear and generalized linear models. Wiley series in probability and statistics.
Antipina, E., Markova, E., & Solodusha, S. (2023). On one approach to computer modeling of the dynamics of heat and power systems by the Volterra series method. In AIP Conference Proceedings, 2757, 060001.
Fomin, O., Polozhaenko, S., Krykun, V., Orlov, A., & Lys, D. (2023). Interpretation of dynamic models based on neural networks in the form of integral-power series. In: Arsenyeva, O., Romanova, T., Sukhonos, M., & Tsegelnyk, Y. (eds.) Smart Technologies in Urban Engineering. STUE 2022. Lecture Notes in Networks and Systems, 536, 258-265, Springer, Cham.
Giannini, F., Colantonio, P., Orengo, G., Serino, A., Stegmayer, G., Pirola, M., & Ghione, G. (2007). Neural networks and Volterra series for time-domain power amplifier behavioral models. International Journal of RF and Microwave Computer-Aided Engineering, 17(2), 160-168.
Khandani, M.K., & Mikhael, W.B. (2020). A study on network size reduction using sparse input representation in time delay neural networks. In 2020 IEEE 63rd International Midwest Symposium on Circuits and Systems (MWSCAS) (pp. 864-867). Springfield, MA, USA.
Liu, W., Na, W., Feng, F., Zhu, L., & Lin, Q. (2020). A Wiener-type dynamic neural network approach to the modeling of nonlinear microwave devices and its applications. In 2020 IEEE MTT-S International Conference on Numerical Electromagnetic and Multiphysics Modeling and Optimization (NEMO) (pp. 1-3). Hangzhou, China.
Liu, W., Na, W., Zhu, L., & Zhang, Q.-J. (2016). A review of neural network based techniques for nonlinear microwave device modeling. In 2016 IEEE MTT-S International Conference on Numerical Electromagnetic and Multiphysics Modeling and Optimization (NEMO) (pp. 1-2). Beijing, China.
Liu, W., Na, W., Zhu, L., Ma, J., & Zhang, Q.-J. (2017). A Wiener-type dynamic neural network approach to the modeling of nonlinear microwave devices. IEEE Transactions on Microwave Theory and Techniques, 65(6), 2043-2062.
Liu, W., Su, Y., and Zhu, L. (2021). Nonlinear device modeling based on dynamic neural networks: A review of methods. In 2021 IEEE 4th International Conference on Electronic Information and Communication Technology (ICEICT) (pp. 662-665). Xi'an, China.
Liu, W., Su, Y., Tan, H., Feng, F., & Zhang, B. (2022). A review of Wiener-type dynamic neural network for nonlinear device modeling. In 2022 IEEE MTT-S International Microwave Workshop Series on Advanced Materials and Processes for RF and THz Applications (IMWSAMP) (pp. 1-3). Guangzhou, China.
Liu, W., Zhu, L., Feng, F., Zhang, W., Zhang, Q.-J., Qian, L., & Liu, G. (2020). A time delay neural network based technique for nonlinear microwave device modelling. Micromachines, 11(9), 831. Basel.
Liu, W., Zhu, L., Na, W., & Zhang, Q.-J. (2016). An overview of neuro-space mapping techniques for microwave device modeling. In 2016 IEEE MTT-S Latin America Microwave Conference (LAMC) (pp. 1-3). Puerto Vallarta, Mexico.
Meruelo, C., Simpson, D.M., Veres, S.M., & Newland, Ph.L. (2016). Improved system identification using artificial neural networks and analysis of individual differences in responses of an identified neuron. Neural Networks, 75, 56-65.
Mitrea, C.A., Lee, C.K.M., & Wu, Z. (2009). A comparison between neural networks and traditional forecasting methods: Case study. International Journal of Engineering Business Management, 1(2), 19-24.
Pavlenko, V.D., & Pavlenko, S.V. (2023). Deterministic identification methods for nonlinear dynamical systems based on the Volterra Model. Applied Aspects of Information Technology, 1(1) (Jun.), 11-32.
Rao, A.R., & Reimherr, M. (2021). Non-linear functional modelling using neural networks. Retrieved from https://arxiv.org/abs/2104.09371.
Rudin, C., & Radin, J. (2019). Why are we using black box models in AI when we don’t need to? A lesson from an explainable AI competition. Harvard Data Science Review, 2(1).
Schoukens, J., & Ljung, L. (2019). Nonlinear system identification: A user-oriented road map. IEEE Control Systems Magazine, 39(6), 28-99.
Sen, J. (2021). Machine Learning – Algorithms, Models and Applications. London, UK, IntechOpen.
Stegmayer, G., Pirola, M., Orengo, G., & Chiotti, O. (2004). Towards a Volterra series representation from a neural network model. WSEAS Transactions on Circuits and Systems, archive 1, 55-61.
Sugiyama, M., Sawai, H., & Waibel, A.H. (1991). Review of TDNN (time delay neural network) architectures for speech recognition. In IEEE International Sympoisum on Circuits and Systems (pp. 582-585), 1.
Todorovic, N., & Klan, P. (2006). State of the art in nonlinear dynamical system identification using artificial neural networks. In 2006 8th Seminar on Neural Network Applications in Electrical Engineering (pp. 103-108). Belgrade, Serbia.
Wang, C.-H., Chen, P.-C., Lin, P.-Z. & Lee, T.-T. (2009). A dynamic neural network model for nonlinear system identification. In 2009 IEEE International Conference on Information Reuse & Integration (pp. 440-441). Las Vegas, NV, USA.
Zhu, L., Zhang, Q., Liu, K., Ma, Y., Peng, B., & Yan, S. (2016). A novel dynamic neurospace mapping approach for nonlinear microwave device modeling. In IEEE Microwave and Wireless Components Letters (pp. 131-133), 26(2).
Downloads
Published
How to Cite
Issue
Section
URN
License
Copyright (c) 2023 Oleksandr Fomin, Viktor Speranskyy, Valentyn Krykun, Oleksii Tataryn, Vladyslav Litynskyi
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
The authors who publish in this journal agree to the following terms:The authors reserve the right to authorship of their work and give the journal the right to first publish this work under the terms of the Creative Commons Attribution License CC BY-NC, which allows other persons to freely distribute published work with a mandatory reference to authors of the original work and the first publication of the work in this journal.
Authors have the right to conclude separate additional agreements for the non-exclusive distribution of the paper in the form in which it was published by this journal (for example, posting work in electronic repository or publishing as part of a monograph), provided that the link to the first publication in this journal is maintained.
The journal policy allows and encourages authors to post on the Internet (for example, in repositories of institutions or on personal websites) the manuscript of work, both before the submission of this manuscript to the editorial staff, and during its editorial work, as it contributes to the emergence of productive scientific discussion and positively affects the efficiency and dynamics of published work citation (see The Effect of Open Access).
