Topology identification of sparse networks of continuous time systems

In this paper, topology identification of dynamic systems in the continuous time (CT) framework is considered. The main objective is to identify the direction of information flow and to estimate the transfer function of each node simultaneously. The output of each node is affected by the outputs of the other nodes along with transportation delay and noise. In contrast to the existing works which are in discrete setting, the current paper uses continuous orthonormal basis to represent the transfer functions. As the network is sparsely connected, the topology identification problem is defined as an estimation of a block-sparse signal from the measurements and a standard lo minimization algorithm known as the block orthogonal matching pursuit (BOMP) is used to identify the topology. The accuracy of the proposed method is demonstrated using data obtained from a system of interconnected tanks. A comparison with discrete time topology identification methods shows that the CT method is more robust and provides the exact topology with higher probability.

[1]  Sami Karjalainen,et al.  Estimating static heat flows in buildings for energy allocation systems , 2006 .

[2]  Joel A. Tropp,et al.  Signal Recovery From Random Measurements Via Orthogonal Matching Pursuit , 2007, IEEE Transactions on Information Theory.

[3]  Yoshito Ohta,et al.  Continuous-time system identification using compactly-supported filter kernels generated from Laguerre basis functions , 2010, 49th IEEE Conference on Decision and Control (CDC).

[4]  Victor Solo,et al.  Topology identification of a sparse dynamic network , 2012, 2012 IEEE 51st IEEE Conference on Decision and Control (CDC).

[5]  R. Tibshirani,et al.  Least angle regression , 2004, math/0406456.

[6]  Mustafa Ayazoglu,et al.  Blind identification of sparse dynamic networks and applications , 2011, IEEE Conference on Decision and Control and European Control Conference.

[7]  M. Eichler.  Granger causality and path diagrams for multivariate time series , 2007 .

[8]  P. Uetz,et al.  Towards an understanding of complex protein networks. , 2001, Trends in cell biology.

[9]  Murti V. Salapaka,et al.  Relations between structure and estimators in networks of dynamical systems , 2011, IEEE Conference on Decision and Control and European Control Conference.

[10]  Donatello Materassi,et al.  Topological identification in networks of dynamical systems , 2008, 2008 47th IEEE Conference on Decision and Control.

[11]  G.J. Pappas,et al.  Identification of stable genetic networks using convex programming , 2008, 2008 American Control Conference.

[12]  Claire J. Tomlin,et al.  Sparse network identifiability via Compressed Sensing , 2016, Autom..

[13]  Albertus C. den Brinker,et al.  Optimality condition for truncated generalized laguerre networks , 1995, Int. J. Circuit Theory Appl..

[14]  A. Papachristodoulou,et al.  Determining Interconnections in Chemical Reaction Networks , 2007, 2007 American Control Conference.

[15]  P. V. D. Hof,et al.  System identification with generalized orthonormal basis functions , 1994, Proceedings of 1994 33rd IEEE Conference on Decision and Control.

[16]  Michael A. Saunders,et al.  Atomic Decomposition by Basis Pursuit , 1998, SIAM J. Sci. Comput..

[17]  Robert D. Nowak,et al.  Causal Network Inference Via Group Sparse Regularization , 2011, IEEE Transactions on Signal Processing.

[18]  C. Granger.  Investigating causal relations by econometric models and cross-spectral methods , 1969 .

[19]  M. Yuan,et al.  Model selection and estimation in regression with grouped variables , 2006 .

[20]  J. Tropp,et al.  SIGNAL RECOVERY FROM PARTIAL INFORMATION VIA ORTHOGONAL MATCHING PURSUIT , 2005 .

[21]  Brett Ninness,et al.  Orthonormal basis functions for continuous-time systems: Completeness and Lp-convergence , 1999, 1999 European Control Conference (ECC).

[22]  Murti V. Salapaka,et al.  On the Problem of Reconstructing an Unknown Topology via Locality Properties of the Wiener Filter , 2010, IEEE Transactions on Automatic Control.

[23]  Lester Melie-García,et al.  Estimating brain functional connectivity with sparse multivariate autoregression , 2005, Philosophical Transactions of the Royal Society B: Biological Sciences.

[24]  James J. Collins,et al.  Inferring Microbial Genetic Networks Statistical learning applied to transcript responses help to gauge how genes influence one another and to identify complex networks , 2004 .

[25]  Michael J. Naylor,et al.  Topology of foreign exchange markets using hierarchical structure methods , 2006, physics/0608084.

[26]  Hugues Garnier,et al.  Continuous-time model identification from sampled data: Implementation issues and performance evaluation , 2003 .

[27]  R. Tibshirani.  Regression Shrinkage and Selection via the Lasso , 1996 .

[28]  Tyrone L. Vincent,et al.  Compressive topology identification of interconnected dynamic systems via Clustered Orthogonal Matching Pursuit , 2011, IEEE Conference on Decision and Control and European Control Conference.

[29]  Arne Dankers,et al.  Errors-in-variables identification in dynamic networks , 2014 .