Recovering the observable part of the initial data of an infinite-dimensional linear system with skew-adjoint generator

Abstract

We consider the problem of recovering the initial data (or initial state) of infinite-dimensional linear systems with unitary semigroups. It is well-known that this inverse problem is well posed if the system is exactly observable, but this assumption may be very restrictive in some applications. In this paper we are interested in systems which are not exactly observable, and in particular, where we cannot expect a full reconstruction. We propose to use the algorithm studied by Ramdani et al. in (Automatica 46:1616-1625 and prove that it always converges towards the observable part of the initial state. We give necessary and sufficient condition to have an exponential rate of convergence. Numerical simulations are presented to illustrate the theoretical results.

Time reversal method with stabilizing boundary conditions for Photoacoustic tomography

“…Appendix) u + is in the energy space (19) and u − is in the corresponding energy space on [−T, 0]. As K ⊂ M int , we see that u belongs to the space (9). Lemma 4 implies that U 0 = 0.…”

“…Let us now consider the general case where u is in (9). Let ψ ∈ C ∞ 0 (− /2, /2), where > 0 is small.…”

Output error minimizing back and forth nudging method for initial state recovery

“…In the latter article the method is called time reversal focusing and they treat the concrete problem of retrieving the initial state of the Kirchhoff plate equation from partial field measurements. Further development of the BFN method includes [10] by Haine showing a partial convergence result when the exact observability assumption is not satisfied, and [9] by Fridman extending the result to a class of semilinear systems. Application to unbounded computational domain is considered by Fliss et al in [8], and a variant for systems containing a diffusive term is suggested by Auroux et al in [4] where the idea is to change the sign of the diffusive term in the backward phase.…”

“…We shall show this only in the skew-adjoint case, but a similar variant for ESAD systems is possible. In addition, we again make the exact observability assumption, but this theorem can be straightforwardly generalized to the non-observable case as is done in [10]. In that case the convergence is not exponential and of course the minimizer is not necessarily unique.…”

New dynamical observer for a batch crystallization process based on solute concentration

“…• If z ∈ ImT ⊕ (ImT ) ⊥ , then the set argmin ψ∈X T ψ − z 2 is closed, convex and non-empty (in particular (8) admits at least one solution).…”