CIBOLA

Synopsis: CIBOLA (Covariance, Incorporateing Basic Option for Linear, Analysis) is a code for relatively rapid modeling of adaptive optics using linear systems methods in the spatial frequency domain (for example, evaluating the performance of an 8 meter class MCAO system takes 1-2 minutes). The distribution includes:

1. cibola.m, cibola_ui.m, cibola_ui.fig — matlab source and figure files
2. example.mat — input and output for a sample MCAO test case
3. cibola_doc.pdf — bare-bones documentation for input and output variables (this may make more sense if you browse the variables in example.mat, with or without the gui, while you read it).
4. maui04.ps — a set of briefing charts providing an overview of the theory used to develop the code.

Known bugs/limitations include:

1. The gui is only usable with Linux.
2. The largest MCAO configuration studied so far has 3 DMs and an 8 meter aperture. Numerical instabilities may appear for systems with more DMs and/or larger apertures.
3. Automated optimization of control loop bandwidths takes longer than necessary (e.g., about 20 times longer than evaluating AO performance for a single, fixed control bandwidth), since Matlab is presently optimizing the control bandwidth to about 1 part in 10⁵ or 10⁶.

Please let me know your experiences and conclusions if you choose to play with this. What's clear, confusing, broken, accurate, inaccurate, or in need of refinement? Possible areas for extensions might include:

• input parameter consistency checking
• graphical output displays
• more flexibility in specifying
• telescope aperture functions and apodization
• optimization fields of view
• wind velocity distributions
• classical servo filters
• ExAO modeling
• nulling WFS spatial aliasing
• Simulations to study PSF speckle statistics
• Estimates for DM dynamic range requirements
• more speculative extensions:
  • Kalman and/or predictive filtering
  • Some sort of model for laser guidestars

Platform: Matlab (Linux required for GUI)

Operating System: Any? (Matlab required. Linux required for GUI.)

Author: Brent Ellerbroek.

Downloadable file format: A single gzip compressed tar file.

Files:	Name			Size		Description
	CIBOLA			-		CIBOLA Software
		cibola.tar.gz		831kB		gziped TAR archive of CIBOLA.

Jump to navigation
Graphical Version

Center for Adaptive Optics