Abstract
The system identification literature has been going through a recent paradigm change with the emergent use of regularization and kernel-based methodologies to identify the process’ impulse response. However, the literature is quite scarce when dealing with processes that possess colored additive output noise. In this case, the current alternative is to identify a system predictor instead, which can be somewhat unfavorable in situations where the process’ model is strictly necessary. So, the main objective of this paper is to introduce a novel regularized system identification methodology that has been specifically developed for the colored output noise scenario. Such methodology is based on the Bayesian perspective of the identification procedure, and it results in the regularized weighted least-squares method, which can be interpreted as an extension of the well-known regularized least squares. The paper also presents the method’s statistical properties, optimal choices, and parametrization structures for both the regularization and weighting matrices, along with a dedicated algorithm to estimate these matrices. Finally, Monte Carlo simulations are performed to demonstrate the method’s efficiency.
Similar content being viewed by others
References
Anderson, B. D. O., & Moore, J. B. (1979). Optimal filtering. Prentice-Hall.
Andersson, C., Wahlström, N., & Schön, T. B. (2018). Data-driven impulse response regularization via deep learning. IFAC-PapersOnLine, 51(15), 1–6. https://doi.org/10.1016/j.ifacol.2018.09.081, 18th IFAC symposium on system identification (SYSID).
Boeira, E., & Eckhard, D. (2021). Regularized impulse response estimation for systems with colored output noise. In: 2021 Australian & New Zealand Control Conference (ANZCC). IEEE, pp. 18–23, https://doi.org/10.1109/ANZCC53563.2021.9628304.
Carlin, B. P., & Louis, T. A. (2000). Bayes and empirical Bayes methods for data analysis (2nd ed.). Chapman & Hall.
Cerone, V., Regruto, D., Abuabiah, M., & Fadda, E. (2018). A kernel-based nonparametric approach to direct data-driven control of LTI systems. IFAC-PapersOnLine, 51(15), 1026–1031. https://doi.org/10.1016/j.ifacol.2018.09.059, 18th IFAC symposium on system identification (SYSID).
Chen, T., & Ljung, L. (2013). Implementation of algorithms for tuning parameters in regularized least-squares problems in system identification. Automatica, 49(7), 2213–2220. https://doi.org/10.1016/j.automatica.2013.03.030
Chen, T., Ohlsson, H., & Ljung, L. (2012). On the estimation of transfer functions, regularizations and gaussian processes—revisited. Automatica, 48(8), 1525–1535. https://doi.org/10.1016/j.automatica.2012.05.026
Cortiella, A., Park, K. C., & Doostan, A. (2021). Sparse identification of nonlinear dynamical systems via reweighted \(\ell _1\)-regularized least squares. Computer Methods in Applied Mechanics and Engineering, 376(113), 620. https://doi.org/10.1016/j.cma.2020.113620
Feng, B., Fu, M., Ma, H., Xia, Y., & Wang, B. (2014). Kalman filter with recursive covariance estimation-sequentially estimating process noise covariance. IEEE Transactions on Industrial Electronics, 61(11), 6253–6263. https://doi.org/10.1109/TIE.2014.2301756
Formentin, S., & Chiuso, A. (2021). Control-oriented regularization for linear system identification. Automatica, 127(109), 539. https://doi.org/10.1016/j.automatica.2021.109539
Formentin, S., & Karimi, A. (2014). Enhancing statistical performance of data-driven controller tuning via \(\ell _2\)-regularization. Automatica, 50(5), 1514–1520. https://doi.org/10.1016/j.automatica.2014.04.001
Forssell, U., & Ljung, L. (2000). Some results on optimal experiment design. Automatica, 36(5), 749–756. https://doi.org/10.1016/S0005-1098(99)00205-8
Fujimoto, Y. (2020). Time-frequency regularization for impulse response estimation. In: 59th annual conference of the society of instrument and control engineers of Japan (SICE). IEEE, pp. 1329–1332, https://doi.org/10.23919/SICE48898.2020.9240275.
Fujimoto, Y. (2021). Kernel regularization in frequency domain: Encoding high-frequency decay property. IEEE Control Systems Letters, 5(1), 367–372. https://doi.org/10.1109/LCSYS.2020.3001879
Gevers, M., Bazanella, A. S., Bombois, X., & Miskovic, L. (2009). Identification and the information matrix: How to get just sufficiently rich? IEEE Transactions on Automatic Control, 54(12), 2828–2840. https://doi.org/10.1109/TAC.2009.2034199
Hjalmarsson, H. (2005). From experiment design to closed-loop control. Automatica, 41(3), 393–438. https://doi.org/10.1016/j.automatica.2004.11.021
Ljung, L. (1999). System identification: theory for the user (2nd ed.). Prentice Hall.
Ljung, L., Singh, R., & Chen, T. (2015). Regularization features in the system identification toolbox. IFAC-PapersOnLine, 48(28), 745–750. https://doi.org/10.1016/j.ifacol.2015.12.219, 17th IFAC symposium on system identification (SYSID).
Ljung, L., Chen, T., & Mu, B. (2019). A shift in paradigm for system identification. International Journal of Control, 93(2), 173–180. https://doi.org/10.1080/00207179.2019.1578407
Matisko, P., & Havlena, V. (2013). Noise covariance estimation for kalman filter tuning using bayesian approach and monte carlo. International Journal of Adaptive Control and Signal Processing, 27(11), 957–973. https://doi.org/10.1002/acs.2369
Mu, B., Chen, T., & Ljung, L. (2017). Tuning of hyperparameters for FIR models-an asymptotic theory. IFAC-PapersOnLine, 50(1), 2818–2823. https://doi.org/10.1016/j.ifacol.2017.08.633, 20th IFAC World Congress.
Mu, B., Chen, T., & Ljung, L. (2018). On asymptotic properties of hyperparameter estimators for kernel-based regularization methods. Automatica, 94, 381–395. https://doi.org/10.1016/j.automatica.2018.04.035
Petersen, K. B., & Pedersen, M. S. (2008). The matrix cookbook. Technical University of Denmark, 7(15), 510.
Pillonetto, G., & De Nicolao, G. (2010). A new kernel-based approach for linear system identification. Automatica, 46(1), 81–93. https://doi.org/10.1016/j.automatica.2009.10.031
Pillonetto, G., Chiuso, A., & De Nicolao, G. (2011). Prediction error identification of linear systems: A nonparametric gaussian regression approach. Automatica, 47(2), 291–305. https://doi.org/10.1016/j.automatica.2010.11.004
Pillonetto, G., Dinuzzo, F., Chen, T., De Nicolao, G., & Ljung, L. (2014). Kernel methods in system identification, machine learning and function estimation: A survey. Automatica, 50(3), 657–682. https://doi.org/10.1016/j.automatica.2014.01.001
Pillonetto, G., Chen, T., Chiuso, A., De Nicolao, G., & Ljung, L. (2022). Regularized system identification: Learning dynamic models from data. Springer International Publishing.
Prando, G., Chiuso, A., & Pillonetto, G. (2017). Maximum entropy vector kernels for MIMO system identification. Automatica, 79, 326–339. https://doi.org/10.1016/j.automatica.2017.01.020
Acknowledgements
This study was financed in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior—Brasil (CAPES)—Finance Code 001, by the Conselho Nacional de Desenvolvimento Científico e Tecnológico—Brasil (CNPq), and by PETROBRAS S/A.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
The authors declare that they have no conflict of interests to influence the work reported in this paper.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendix A Matrix Calculus Results
Appendix A Matrix Calculus Results
Here we present some matrix calculus results that were used to calculate the results in Sect. 3 regarding the MSE trace optimization for the RWLS estimator.
1.1 A.1 Derivatives with Relation to P
Let us start with some generic derivatives that were used to compute the derivative of the MSE trace with relation to P:
where \(Z=\left( A^TP^T+I\right) ^{-1}\), and \(W=\left( BP^T+I\right) ^{-1}\).
1.2 A.2 Derivatives with Relation to M
Now, some results regarding the derivatives of the MSE trace with relation to M:
with \(V=\left( B^TM^TA^T+I\right) ^{-1}\), and \(U=\left( BM^TC+I\right) ^{-1}\).
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
Boeira, E., Eckhard, D. The Optimal Regularized Weighted Least-Squares Method for Impulse Response Estimation. J Control Autom Electr Syst 34, 302–314 (2023). https://doi.org/10.1007/s40313-022-00968-1
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s40313-022-00968-1