Abstract
This study proposes a decomposition optimization-based expectation maximization algorithm for switching models. The identities of each sub-model are estimated in the expectation step, while the parameters are updated using the decomposition optimization method in the maximization step. Compared with the traditional expectation maximization algorithm and the gradient descent expectation maximization algorithm, the decomposition optimization-based expectation maximization algorithm avoids the matrix inversion and eigenvalue calculation; thus, it can be extended to complex nonlinear models and large-scale models. Convergence analysis and simulation examples are given to show the effectiveness of the proposed algorithm.
Similar content being viewed by others
Data Availability
The data used to support the findings of this study are available from the corresponding author upon request.
References
Yan, H.Y., Zhu, Y.G.: Bang-bang control model for uncertain switched systems. Appl. Math. Model. 39(10–11), 2994–3002 (2015)
Wu, F.Y., Lian, J.: Stabilization of constrained switched systems via multiple Lyapunov R-functions. Syst. Control Lett. (2020). https://doi.org/10.1016/j.sysconle.2020.104686
Zhou, Y.H., Zhang, X.: Partially-coupled nonlinear parameter optimization algorithm for a class of multivariate hybrid models. Appl. Math. Comput. 414, 126663 (2022)
Liu, T., Huang, J.: Discrete-time distributed observers over jointly connected switching networks and an application. IEEE Trans. Autom. Control 66(4), 1918–1924 (2020)
Caswell, H.: Construction, Analysis, and Interpretation. Sinauer, Sunderland (2001)
Navarro, G.S., Gallegos, J.A.: On the property sign-stability of equilibria in quasimonotone positive nonlinear systems. In: Proc of the 33rd IEEE Conf Decis Control., vol. 4, pp. 4043–4048 (1994)
Shorten, R., Wirth, F., Leith, D.: A positive systems model of TCP-like congestion control: asymptotic results. IEEE/ACM Trans. Netw. 14(3), 616–629 (2006)
Vidal, R.: Recursive identification of switched ARX systems. Automatica 44(2), 2274–2287 (2008)
Wang, H., Fan, H., Pan, J.: Complex dynamics of a four-dimensional circuit system. Int. J. Bifurc. Chaos 31(14), 2150208 (2021)
Li, M.H., Liu, X.M., et al.: Least-squares-based iterative and gradient-based iterative estimation algorithms for bilinear systems. Nonlinear Dyn. 89(1), 197–211 (2017)
Zhou, Y.H., Zhang, X.: Hierarchical estimation approach for RBF-AR models with regression weights based on the increasing data length. IEEE Trans. Circuits Syst. II Express Briefs 68(12), 3597–3601 (2021)
Li, J.M., Ding, F.: Identification methods of nonlinear systems based on the kernel functions. Nonlinear Dyn. 104(3), 2537–2552 (2021)
Xu, L., Ding, F., Zhu, Q.: Separable synchronous multi-innovation gradient-based iterative signal modeling from on-line measurements. IEEE Trans. Instrum. Meas. 71, 6501313 (2022)
Söderström, T., Soverini, U.: Errors-in-variables identification using maximum likelihood estimation in the frequency domain. Automatica 79, 131–143 (2017)
Garulli, A., Paoletti, S., Vicino, A.: A survey on switched and piecewise affine system identification. IFAC Symp. Syst. Ident. 45(16), 344–355 (2012)
Bianchi, F., Breschi, V., Piga, D., Piroddi, L.: Model structure selection for switched NARX system identification: a randomized approach. Automatica 125, 109415 (2021)
Lauer, F., Bloch, G.: Hybrid System Identification. Springer, Berlin (2019)
Ma, Y.J., Zhao, S.Y., Huang, B.: Multiple-model state estimation based on variational Bayesian inference. IEEE Trans. Autom. Control 64(4), 1679–1685 (2019)
Moon, T.K.: The expectation-maximization algorithm. IEEE Signal Process. Mag. 13(6), 47–60 (1996)
Wang, D.Q., Zhang, S., et al.: A novel EM identification method for Hammerstein systems with missing output data. IEEE Trans. Ind. Inform. 16(4), 2500–2508 (2020)
Chen, J., Huang, B., et al.: Variational Bayesian approach for ARX systems with missing observations and varying time-delays. Automatica 94, 194–204 (2018)
Ma, J.X., Huang, B., et al.: Iterative identification of Hammerstein parameter varying systems with parameter uncertainties based on the variational Bayesian approach. IEEE Trans. Syst. Man Cyber. Syst. 50(3), 1035–1045 (2020)
Lu, Y.J., Huang, B., Khatibisepehr, S.: A variational Bayesian approach to robust identification of switched ARX models. IEEE Trans. Cyber. 46(12), 3195–3208 (2016)
Yang, X.Q., Yin, S.: Robust global identification and output estimation for LPV dual-rate systems subjected to random output time-delays. IEEE Trans. Ind. Inform. 13(6), 2876–2885 (2017)
Xu, L.: Separable multi-innovation Newton iterative modeling algorithm for multi-frequency signals based on the sliding measurement window. Circuits Syst. Signal Process. 41(2), 805–830 (2022)
Gan, M., Chen, X.X., et al.: Adaptive RBF-AR models based on multi-innovation least squares method. IEEE Signal Process. Lett. 26(8), 1182–1186 (2019)
Xu, H., Ding, F., Yang, E.F.: Modeling a nonlinear process using the exponential autoregressive time series model. Nonlinear Dyn. 95(3), 2079–2092 (2019)
Chen, J., Zhu, Q.M., Hu, M.F., Guo, L.X., Narayan, P.: Improved gradient descent algorithms for time-delay rational state-space systems: intelligent search method and momentum method. Nonlinear Dyn. 101(1), 361–373 (2020)
Xu, L., Yang, E.F.: Auxiliary model multiinnovation stochastic gradient parameter estimation methods for nonlinear sandwich systems. Int. J. Robust Nonliear Control 31(1), 148–165 (2021)
Wang, D.Q.: Hierarchical parameter estimation for a class of MIMO Hammerstein systems based on the reframed models. Appl. Math. Lett. 57, 13–19 (2016)
Wang, D.Q., Mao, L., et al.: Recasted models based hierarchical extended stochastic gradient method for MIMO nonlinear systems. IET Control Theory Appl. 11(4), 476–485 (2017)
Ding, F., Zhang, X., Xu, L.: The innovation algorithms for multivariable state-space models. Int. J. Adapt. Control Signal Process. 33(11), 1601–1608 (2019)
Bai, E.W.: Identification of linear systems with hard input nonlinearities of known structure. Automatica 38(5), 853–860 (2002)
Ueda, N., Nakano, R.: Deterministic annealing EM algorithm. Neural Netw. 11, 271–282 (1998)
Gan, M., Chen, G.Y., et al.: Term selection for a class of nonlinear separable models. IEEE Trans. Neural Netw. Learn. Syst. 31(2), 445–451 (2020)
Zhou, Y.H., Ding, F.: Modeling nonlinear processes using the radial basis function-based state-dependent autoregressive models. IEEE Signal Process. Lett. 27, 1600–1604 (2020)
Hou, J., Chen, F.W., Zhu, Z.Q.: Gray-box parsimonious subspace identification of Hammerstein-type systems. IEEE Trans. Ind. Electron. 68(10), 9941–9951 (2021)
Chen, J., Zhu, Q.M., et al.: Interval error correction auxiliary model based gradient iterative algorithms for multi-rate ARX models. IEEE Trans. Autom. Control 65(10), 4385–4392 (2020)
Zhang, X.: Optimal adaptive filtering algorithm by using the fractional-order derivative. IEEE Signal Process Lett. 29, 399–403 (2022)
Diederik, P.K., Jimmy, L.B.: ADAM: A method for stochastic optimization. In: Int. Conf. Lear. Represent, San Diego, USA, pp. 7–9 (2015)
Ruder, S.: An overview of gradient descent optimization algorithms. arXiv:1609.04747v2 [cs.LG] (2017)
Duchi, J., Hazan, E., Singer, Y.: Adaptive subgradient methods for online learning and stochastic optimization. J. Mach. Learn. Res. 12, 2121–2159 (2011)
Bako, L., Boukharouba, K., Duviella, E., Lecoeuche, S.: A recursive identification algorithm for switched linear/affine models. Nonlinear Anal. Hybrid Syst. 5, 242–253 (2011)
Aggoune, L., Chetouani, Y., Raïsi, T.: Fault detection in the distillation column process using Kullback Leibler divergence. ISA Trans. 63, 394–400 (2016)
Acknowledgements
The authors would like to thank the associate editor and the anonymous reviewers for their constructive and helpful comments and suggestions to improve the quality of this paper.
Funding
This work is supported by the National Natural Science Foundation of China (No. 61973137), the Natural Science Foundation of Jiangsu Province (No. BK20201339) and the Funds of the Science and Technology on Near-Surface Detection Laboratory (Nos. 61424140207, 61424140202).
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
The authors declare that they have no conflict of interest.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
This work is supported by the National Natural Science Foundation of China (No. 61973137), the Natural Science Foundation of Jiangsu Province (No. BK20201339) and the Funds of the Science and Technology on Near-Surface Detection Laboratory (Nos. 61424140207, 61424140202).
Appendix
Appendix
Proof of of Theorem 1
In iteration \(k-1\), the Q function can be written by
According to Jensen’s inequality, the right side of the above equation satisfies
Based on the Kullback–Leibler divergence [44], when
the Q function is
that is
When \(p^k_p(I(t)=i)\) has been obtained, the parameter estimates \({\textbf{W}}_{k}\) of the LS algorithm satisfy
while the estimates \({\textbf{W}}_{k}\) by using the GD-EM and DO-EM (AA-DO-EM) algorithms can guarantee
Therefore, it follows that
\(\square \)
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Chen, J., Mao, Y., Hu, M. et al. Decomposition optimization method for switching models using EM algorithm. Nonlinear Dyn 111, 9361–9375 (2023). https://doi.org/10.1007/s11071-023-08302-3
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11071-023-08302-3