Modal control is an established tool in adaptive optics. It allows not only for the reduction in the controllable degrees of freedom, but also for filtering out unseen modes and optimizing gain on a mode-by-mode basis. When Zernike polynomials are employed as the modal basis for correcting atmospheric turbulence, their cross-correlations translate to correction errors. We propose optimal modal decomposition for gradient-descent-based wavefront sensorless adaptive optics, which is free of this problem. We adopt statistically independent Karhunen-Loève functions for iterative blind correction and analyze performance of the algorithm in static as well as in dynamic simulated turbulence conditions.

中文翻译：

---

大气湍流中的最优盲搜索模态波前校正。第一部分：模拟

模态控制是自适应光学中已建立的工具。它不仅允许减小可控制的自由度，而且还可以滤除看不见的模式并在逐个模式的基础上优化增益。当将Zernike多项式用作校正大气湍流的模态基础时，它们的互相关会转换为校正误差。我们提出了基于梯度下降的波前无传感器自适应光学器件的最佳模态分解，它没有这个问题。我们采用统计独立的Karhunen-Loève函数进行迭代盲校正，并分析算法在静态和动态模拟湍流条件下的性能。