Keywords: 
• 变形镜高斯函数指数 /
• 迭代矩阵 /
• 斜率响应矩阵 /
• 稀疏度

Abstract: Among all kinds of wavefront reconstruction algorithms in adaptive optical systems, the standard and mostly used algorithm is the direct gradient wavefront reconstruction algorithm. As the number of sub-apertures in Shack-Hartmann wavefront sensor and the actuators for deformable mirror increases, the reconstruction matrix in direct gradient wavefront reconstruction algorithm takes too much space and the number of multiplication in the algorithm increases sharply. So, the iterative algorithm is adopted in wavefront reconstruction for the high-resolution adaptive optical system. The number of multiplication and the required space of the iterative algorithm are directly related to the sparseness of both iterative matrix and slope response matrix. In an adaptive optical system, the sparseness of these two matrixes is connected with the system parameters. Therefore, it is necessary to study how to choose the proper parameters for an adaptive optical system when it uses iterative wavefront reconstruction algorithm. In this paper, the sparseness of slope response matrix and iterative matrix are analyzed based on a 613-actuator adaptive optical system. The influence of the Gaussian function index of deformable mirror on the sparsenesses of slope response matrix, iterative matrix, stability and correction qualities of the adaptive optical system are also studied under the condition of constant actuator spacing and coupling coe?cient. A larger Gaussian function index results in a lower sparseness of the slope response matrix and the iterative matrix. Too large or too small a Gaussian function index will degrade the stability and the correction quality of an adaptive optical system. Finally, the optimal range of the Gaussian function index is provided by balancing the sparseness of slope response matrix, the correction quality, and the stability of the adaptive optical system.