In this paper we present two algorithms for estimating the aero-optical aberration of a transonic flow around a 2-D turret based on Malley probe signals or pressure signals from a few selected points. These two algorithms use Artificial Neural Networks and Linear Stochastic Estimation of varying model orders to estimate Proper Orthogonal Decomposition modal coefficients. These estimated coefficients are then used to reconstruct an estimated wavefront. This estimated wavefront is subtracted from the true wavefront to obtain a simulated reduction in the overall level of optical aberration. Reductions of up to 48% are achieved for both models. A robustness analysis is also performed, in which it is found that the algorithm is not sufficiently robust to changing flow conditions. Solutions are proposed for further investigation.
展开▼
