Sensor Selection in Presence of Random Failures

We analyze the observability of a dynamical system when each sensor is subject to random failure. In particular, we model the fact that each sensor may fail independently of the other and this failure is assumed to be a Bernoulli random variable with known parameter. We leverage results from random matrix theory to obtain probabilistic bounds on three metrics of observability, viz. the (negative of) maximum eigenvalue, the minimum eigenvalue and the (negative of) trace of the inverse of the observability Gramian. The goal is to perform sensor selection to maximize the expected value of the metric, which we show becomes equivalent to optimizing the metric evaluated over the expected value of the Gramian, with a known probability. A greedy algorithm is then used to perform the selection for which we characterize the factor of sub-optimality relative to the optimal corresponding to each metric. For a specific class of systems, the analytic bounds are reasonable for the extreme eigenvalues, but can be very conservative for the trace of the inverse Gramian.

[1]  Maryam Kamgarpour,et al.  Performance guarantees for greedy maximization of non-submodular set functions in systems and control , 2017, ArXiv.

[2]  Soummya Kar,et al.  Minimum cost input/output design for large-scale linear structural systems , 2014, Autom..

[3]  Ufuk Topcu,et al.  A Randomized Greedy Algorithm for Near-Optimal Sensor Scheduling in Large-Scale Sensor Networks , 2017, 2018 Annual American Control Conference (ACC).

[4]  John Lygeros,et al.  On Submodularity and Controllability in Complex Dynamical Networks , 2014, IEEE Transactions on Control of Network Systems.

[5]  Marc M. J. van de Wal,et al.  A review of methods for input/output selection , 2001, Autom..

[6]  H. Weber,et al.  Analysis and optimization of certain qualities of controllability and observability for linear dynamical systems , 1972 .

[7]  George J. Pappas,et al.  Sensor placement for optimal Kalman filtering: Fundamental limits, submodularity, and algorithms , 2015, 2016 American Control Conference (ACC).

[8]  Stephen P. Boyd,et al.  Sensor Selection via Convex Optimization , 2009, IEEE Transactions on Signal Processing.

[9]  Alexander Olshevsky,et al.  Minimal Controllability Problems , 2013, IEEE Transactions on Control of Network Systems.

[10]  Alex Gittens,et al.  TAIL BOUNDS FOR ALL EIGENVALUES OF A SUM OF RANDOM MATRICES , 2011, 1104.4513.

[11]  Shreyas Sundaram,et al.  Sensor selection for Kalman filtering of linear dynamical systems: Complexity, limitations and greedy algorithms , 2017, Autom..

[12]  Ramesh Govindan,et al.  Utility based sensor selection , 2006, IPSN.

[13]  Andreas Krause,et al.  Guarantees for Greedy Maximization of Non-submodular Functions with Applications , 2017, ICML.

[14]  Age K. Smilde,et al.  Selection of optimal sensor position in a tubular reactor using robust degree of observability criteria , 2000 .

[15]  Stephen L. Smith,et al.  Submodularity and greedy algorithms in sensor scheduling for linear dynamical systems , 2015, Autom..

[16]  Joel A. Tropp,et al.  User-Friendly Tail Bounds for Sums of Random Matrices , 2010, Found. Comput. Math..

[17]  Francesco Bullo,et al.  Controllability Metrics, Limitations and Algorithms for Complex Networks , 2013, IEEE Transactions on Control of Network Systems.