借助影响函数矩阵,分析了变形镜拟合波前像差的性能。提出了一种滤除高阶像差模式影响的波前控制算法,弥补了最速下降法无法通过模式选择优化校正过程的缺点,提高了变形镜的校正能力。对人眼出射的畸变波前进行实验,结果表明,经过6次迭代以后,波前的均方根值(RMS)达到衍射极限,系统闭环校正频率为15 Hz。说明基于该控制算法的微机械薄膜变形镜自适应光学系统能够实时校正动态人眼像差,为搭建小型化、低成本的人眼波前像差校正系统提供了算法支持。

The deformable mirror fitting ability of wavefront aberration is analyzed through influence function matrix. A wavefront control algorithm which filters the influence of high-order mode aberration is proposed. It makes up the deficiency of the steepest descent method which cannot optimize the correction process by selecting aberration mode,and improves the correction ability of deformable mirror. Wavefront distortion of a human eye is used in the experiment as the incident wavefront. The results show that the root-mean-square (RMS) value of wavefront aberration can decline to the diffraction limit after six iterations and the frequency of the closed-loop calibration system is 15 Hz. This indicates that micromachined membrane deformable mirror adaptive optics system based on the control algorithm can meet the demand of real-time human eye aberration correction. It offers an appropriate algorithm for small and low-cost system of human eye wavefront aberration correction.