For adaptive optical systems to compensate for atmospheric turbulence
effects, the wave-front perturbation must be measured with a wave-fron
t sensor (WFS) and corrected with a deformable mirror. One limitation
in this process is the time delay between the measurement of the aberr
ated wave front and implementation of the proper correction. Statistic
al techniques exist for predicting the atmospheric aberrations at the
time of correction based on the present and past measured wave fronts.
However, for the statistical techniques to be effective, key paramete
rs of the atmosphere and the adaptive optical system must be known. Th
ese parameters include the Fried coherence length r(0), the atmospheri
c wind-speed profile, and the WFS slope measurement error. Neural netw
orks provide nonlinear solutions to adaptive optical problems while of
fering the possibility to function under changing seeing conditions wi
thout actual knowledge of the current state of the key parameters. We
address the use of neural networks for WFS slope measurement predictio
n with only the noisy WFS measurements as inputs. Where appropriate, w
e compare with classical statistical-based methods to determine if neu
ral networks offer true benefits in performance. (C) 1997 Optical Soci
ety of America