Control system design for retinal imaging adaptive optics systems

In this paper, the control problem in retinal imaging adaptive optics systems is generalized to that of shape control for a flexible membrane representing a deformable membrane mirror. Due to the dynamic nature of the aberrations in the eye, the shape control problem addressed is the tracking of an unknown and time-varying shape for a distributed membrane (i.e. desired shape of the mirror). The proposed controller design approach relies on constructing a Q-parameterized set of stabilizing controllers for the system under consideration and the online tuning of the Q-parameter so that the controller converges to the controller needed to achieve regulation. Partial decoupling of the multi-input multi-output closed loop system dynamics is introduced to allow the tuning to be performed based on decentralized adaptation algorithms.