Controllers for Adaptive Optics Systems using Formal Synthesis

Professor Doug Looze

University of Massachusetts

Department of Electrical and Computing Engineering

This talk presents a formulation of the adaptive optics minimum variance control problem as a discrete-time optimization. The formulation incorporates the wavefront sensor frame integration in discrete-time models of the deformable mirror and incident wavefront. The formulation results in a Linear-Quadratic-Gaussian (LQG) problem. The structure of the LQG solution under ideal conditions is related to the standard adaptive optics integral controller. Although the optimization is a Linear-Quadratic-Gaussian problem, the problem is also solved using H∞ synthesis. The LQG, H∞, and standard integral controllers are compared using two techniques. Simulation comparisons use a CAOS simulation that represents the adaptive optics system with a pyramid sensor at the Calar Alto Observatory near Almeria, Spain. The controllers are also compared using tip-tilt data from the Keck I Observatory in Hawaii.

Friday, March 17, 2006

3:30 – 4:30 p.m.

1500 EECS