Distributed Kalman filtering compared to Fourier domain preconditioned conjugate gradient for laser guide star...

Area of Science:

• Astronomy and Astrophysics
• Optical Engineering
• Computational Science

Background:

• Adaptive optics (AO) systems are crucial for high-resolution astronomical imaging.
• Laser guide star (LGS) AO requires efficient wavefront reconstruction algorithms to correct atmospheric turbulence.
• Existing methods like Fourier domain preconditioned conjugate gradient (FDPCG) have limitations in performance and computational efficiency.

Purpose of the Study:

• To compare the performance and computational cost of two Fourier-based tomographic wavefront reconstruction algorithms for LGS AO.
• To evaluate the distributed Kalman filter (DKF) for its suitability in real-time AO systems, specifically for the Thirty Meter Telescope.
• To assess the sensitivity of the DKF algorithm to uncertainties in wind profile prior information.

Main Methods:

• Implementation and analysis of the iterative FDPCG algorithm with pseudo-open-loop control (POLC).
• Implementation and analysis of the noniterative, spatially invariant DKF algorithm in the Fourier domain.
• Performance and cost analysis for a LGS multiconjugate AO system, including DKF's sensitivity to wind profile errors.

Main Results:

• Both FDPCG and DKF algorithms have a computational cost proportional to N log(N).
• DKF demonstrates significantly reduced wavefront error compared to FDPCG when wind profile accuracy is within 10% speed and 20° direction.
• DKF's nonsequential nature and parallelism make it suitable for real-time GPU implementation.

Conclusions:

• The DKF algorithm provides a more effective wavefront reconstruction solution for LGS AO systems than FDPCG.
• DKF's robustness to wind profile uncertainties and its suitability for GPU acceleration are key advantages for next-generation telescopes.
• DKF represents a promising advancement for real-time wavefront control in large astronomical AO systems.