Title :
Fast finite-horizon kalman filter in wavefront estimation for adaptive optics
Author :
Massioni, Paolo ; Di Loreto, Michael
Author_Institution :
INSA de Lyon, Univ. de Lyon, Villeurbanne, France
Abstract :
We consider the problem of wavefront estimation for the control of adaptive optics systems. It has been shown that an improvement in the performance can be achieved by assuming a turbulence dynamical model which allows the use of a Kalman filter; nevertheless, the implementation of such filters has proven to be difficult in real-time large-scale systems due to memory, computational costs and latency issues. In this article we propose a possible work around of such problem, which consists in using the non-recursive expression of a finite-horizon Kalman filter, which only requires a matrix-vector multiplication at run-time. This allows the use of complex state-space models for the estimation, without the need of explicitly computing the state vector at each time step, yielding a significant gain in term of computational time.
Keywords :
Kalman filters; adaptive control; adaptive optics; large-scale systems; matrix algebra; optical control; optimal control; state-space methods; Earth based telescope; adaptive optics system; fast finite-horizon Kalman filter; matrix-vector multiplication; real-time large-scale system; turbulence dynamical model; wavefront estimation; Adaptive optics; Computational efficiency; Estimation; Kalman filters; Telescopes; Time measurement; Vectors;
Conference_Titel :
Control Conference (ECC), 2014 European
Conference_Location :
Strasbourg
Print_ISBN :
978-3-9524269-1-3
DOI :
10.1109/ECC.2014.6862311