Design of a Linear Gaussian Control Law for an Adaptive Optics System
Abstract
This thesis considers the design of a linear quadratic Gaussian (LQG) controller for a ground-based adaptive-optics telescopes. The incoming aberrated image is reflected from a 97-element piezoelectric mirror, then measured with a Hartmann-type wavefront sensor. A Kalman filter processes the outputs of the wavefront sensor and obtains estimates of system states. A linear quadratic regulator processes these state estimates and determines an appropriate set of commands for the deformable mirror. Atmospheric distortion is modeled as a set of fourteen Zernike coefficients whose dynamic behavior is produced by excitation of a set of shaping filters by zero-mean Gaussian white noise. The response of the mirror to control voltages is modeled as a set of Zernike coefficients whose dynamics are modeled as deterministic first-order systems. The entire control system is simulated using the Multimode Simulation for Optimal Filter Evaluation (MSOFE) software.
Document Details
- Document Type
- Technical Report
- Publication Date
- Dec 01, 1990
- Accession Number
- ADA230547
Entities
People
- Mark A. Von Bokern
Organizations
- Air Force Institute of Technology