A laboratory demonstration of an LQG technique for correcting frozen flow turbulence in adaptive optics systems

TL;DR

This paper demonstrates a laboratory implementation of an LQG-based predictive control method to mitigate frozen-flow atmospheric turbulence in adaptive optics, resulting in significant wavefront error reduction and improved image quality.

Contribution

It introduces 'Predictive Fourier Control', a novel LQG technique for real-time correction of frozen-flow turbulence in adaptive optics systems.

Findings

Threefold reduction in residual wavefront error

10% increase in Strehl ratio

Effective real-time turbulence prediction

Abstract

We present the laboratory verification of a method for re- moving the effects of frozen-flow atmospheric turbulence using a Linear Quadratic Gaussian (LQG) controller, also known as a Kalman Filter. This method, which we term "Predictive Fourier Control," can identify correlated atmospheric motions due to layers of frozen flow turbulence, and can predictively remove the effects of these correlated motions in real-time. Our laboratory verification suggests a factor of 3 improvement in the RMS residual wavefront error and a 10% improvement in measured Strehl of the system. We found that the RMS residual wavefront error was suppressed from 35.0 nm to 11.2 nm due to the use of Predictive Fourier Control, and that the far field Strehl improved from 0.479 to 0.520.