Non-Asymptotic Kernel-Based Parametric Estimation of Continuous-Time Linear Systems

In this paper, a novel framework to address the problem of parametric estimation for continuous-time linear time-invariant dynamic systems is dealt with. The proposed methodology entails the design of suitable kernels of non-anticipative linear integral operators thus obtaining estimators showing, in the ideal case, “non-asymptotic” (i.e., “finite-time”) convergence. The analysis of the properties of the kernels guaranteeing such a convergence behaviour is addressed and a novel class of admissible kernel functions is introduced. The operators induced by the proposed kernels admit implementable (i.e., finite-dimensional and internally stable) state-space realizations. Extensive numerical results are reported to show the effectiveness of the proposed methodology. Comparisons with some existing continuous-time estimators are addressed as well and insights on the possible bias affecting the estimates are provided.

[1]  Peter C. Young,et al.  Recursive Estimation and Time-Series Analysis: An Introduction , 1984 .

[2]  Torsten Söderström,et al.  Editorial special issue on continuous-time model identification , 2011 .

[3]  S. Sastry,et al.  Adaptive Control: Stability, Convergence and Robustness , 1989 .

[4]  Marion Gilson,et al.  Developments for the matlab contsid toolbox , 2003 .

[5]  Lennart Ljung,et al.  Initialisation aspects for subspace and output-error identification methods , 2003, 2003 European Control Conference (ECC).

[6]  Petre Stoica,et al.  Decentralized Control , 2018, The Control Systems Handbook.

[7]  Hugues Garnier,et al.  Numerical illustrations of the relevance of direct continuous-time model identification , 2002 .

[8]  R. Brockett.  Finite Dimensional Linear Systems , 2015 .

[9]  T. A. Burton,et al.  Volterra integral and differential equations , 1983 .

[10]  P. Young,et al.  Refined instrumental variable methods of recursive time-series analysis Part III. Extensions , 1980 .

[11]  P. Young.  Unified estimation of discrete and continuous-time transfer function models , 2008 .

[12]  Hugues Garnier,et al.  The CONTSID Toolbox: A Matlab Toolbox for Continuous-Time System Identification , 2000 .

[13]  Peter C. Young,et al.  Refined instrumental variable methods for identification of Hammerstein continuous-time Box-Jenkins models , 2008, 2008 47th IEEE Conference on Decision and Control.

[14]  Liuping Wang,et al.  Identification of Continuous-time Models from Sampled Data , 2008 .

[15]  Lennart Ljung,et al.  System Identification: Theory for the User , 1987 .

[16]  L. Shaw,et al.  System Theory. , 1965, Science.

[17]  M. Lovera,et al.  Continuous-time predictor-based subspace identification using laguerre filters , 2011 .

[18]  W. Marsden.  I and J , 2012 .

[19]  Lennart Ljung,et al.  Issues in sampling and estimating continuous-time models with stochastic disturbances , 2010, Autom..