State covariances and the matrix completion problem

State statistics of a linear system obey certain structural constraints that arise from the underlying dynamics and the directionality of input disturbances. Herein, we formulate completion problems of partially known state statistics with the added freedom of identifying disturbance dynamics. The goal of the proposed completion problem is to obtain information about input excitations that explain observed sample statistics. Our formulation aims at low-complexity models for admissible disturbances. The complexity represents the dimensionality of the subspace of the state-dynamics that is directly affected by disturbances. An example is provided to illustrate that colored-in-time stochastic processes can be effectively used to explain available data.

[1]  Robert E. Mahony,et al.  The geometry of weighted low-rank approximations , 2003, IEEE Trans. Signal Process..

[2]  Tommi S. Jaakkola,et al.  Weighted Low-Rank Approximations , 2003, ICML.

[3]  Emmanuel J. Candès,et al.  Exact Matrix Completion via Convex Optimization , 2008, Found. Comput. Math..

[4]  Rita Simões,et al.  On the Lyapunov and Stein equations , 2007 .

[5]  Tryphon T. Georgiou,et al.  An ADMM algorithm for matrix completion of partially known state covariances , 2013, 52nd IEEE Conference on Decision and Control.

[6]  Ruslan Salakhutdinov,et al.  Collaborative Filtering in a Non-Uniform World: Learning with the Weighted Trace Norm , 2010, NIPS.

[7]  Charles R. Johnson,et al.  Matrix analysis , 1985, Statistical Inference for Engineers and Data Scientists.

[8]  Tryphon T. Georgiou,et al.  The structure of state covariances and its relation to the power spectrum of the input , 2002, IEEE Trans. Autom. Control..

[9]  Bassam Bamieh,et al.  Componentwise energy amplification in channel flows , 2005, Journal of Fluid Mechanics.

[10]  Stephen P. Boyd,et al.  A rank minimization heuristic with application to minimum order system approximation , 2001, Proceedings of the 2001 American Control Conference. (Cat. No.01CH37148).

[11]  Tryphon T. Georgiou,et al.  Spectral analysis based on the state covariance: the maximum entropy spectrum and linear fractional parametrization , 2002, IEEE Trans. Autom. Control..

[12]  Stephen P. Boyd,et al.  Convex Optimization , 2004, Algorithms and Theory of Computation Handbook.

[13]  Pablo A. Parrilo,et al.  Guaranteed Minimum-Rank Solutions of Linear Matrix Equations via Nuclear Norm Minimization , 2007, SIAM Rev..

[14]  Charles R. Johnson,et al.  Possible inertia combinations in the Stein and Lyapunov equations , 1995 .

[15]  Lieven Vandenberghe,et al.  Interior-Point Method for Nuclear Norm Approximation with Application to System Identification , 2009, SIAM J. Matrix Anal. Appl..

[16]  Andrea Montanari,et al.  Matrix completion from a few entries , 2009, 2009 IEEE International Symposium on Information Theory.

[17]  Emmanuel J. Candès,et al.  Matrix Completion With Noise , 2009, Proceedings of the IEEE.

[18]  Rashad Moarref,et al.  Model-based design of transverse wall oscillations for turbulent drag reduction , 2012, Journal of Fluid Mechanics.

[19]  Emmanuel J. Candès,et al.  The Power of Convex Relaxation: Near-Optimal Matrix Completion , 2009, IEEE Transactions on Information Theory.

[20]  M. Dahleh,et al.  Energy amplification in channel flows with stochastic excitation , 2001 .

[21]  Stephen P. Boyd,et al.  Distributed Optimization and Statistical Learning via the Alternating Direction Method of Multipliers , 2011, Found. Trends Mach. Learn..

[22]  P. Ioannou,et al.  Stochastic forcing of the linearized Navier–Stokes equations , 1993 .

[23]  G. F. Bryant.  On a generalization of a theorem of Lyapunov , 1975 .