Search

Abstract: The adaptive optics system (AOS) often operates in a discrete sampling process with finite closed-loop frequency. Reconstruction, detection, and time lag induced errors are the main correction errors of the system. An AOS that is based on a liquid crystal (LC) benefits from the LC’s high correction precision, thus the reconstruction error can be ignored. The primary error will be induced by the time lag from the time of detection to the time of compensation. In this paper, some theoretical simulations are introduced in order to evaluate the correction precision of AOS with an LC corrector. The main purpose is to compare the correction precision between the open-loop and closed-loop control. We attempt to find a method to ascertain the exact precision of the open-loop control and show whether it improves the correction precision. The conclusion is thus reached that the actual error rejection bandwidth for the closed-loop was lower than the −3dB error rejection bandwidth measured in practice. The increased refresh frequency of the open-loop control can improve the imaging performance to nearly −3dB bandwidth of the detector measured, which is the maximum possible bandwidth due to the time lag. [Copyright &y& Elsevier]

Abstract The performance of free-space laser communication (FSLC) is greatly decreased by horizontal atmospheric turbulence. Adaptive optics is used to overcome horizontal turbulence. This study investigated the effect of the 3 dB bandwidth of an adaptive optics system (AOS) on FSLC performance by using power-in-bucket (PIB) method, which considers the effect of receiving aperture and then, can accurately evaluate of the effect of system bandwidth on the communication performance. A 9 km horizontal maritime link experiment was performed through adaptive correction to validate the theoretical analysis. Results indicate that the effect of system bandwidth can be accurately evaluated using PIB method. This work provides a basis for designing and evaluating AOSs for FSLC. [ABSTRACT FROM AUTHOR]

Abstract: To process real-time wavefront signal in our liquid crystal adaptive optics system (LCAOS), a real-time wavefront processor (RTWP) with graphics processing unit (GPU) accelerator is developed. A series of parallel acceleration algorithm is implemented to reconstruct wavefront, compute gray map and process lookup table (LUT). Especially, an algorithm by using symmetry of Zernike polynomial is designed to compute gray map and it provides a speedup of 3.16. Finally, the total computing time in our RTWP is 109μs and temporal bandwidth in our LCAOS is 45Hz; it is shown that our RTWP has the ability to real time process wavefront signal in our LCAOS. [Copyright &y& Elsevier]