基于扩展卡尔曼滤波的高对比度成像星光和偏差递推估计
Recursive Starlight and Bias Estimation for High-Contrast Imaging with an Extended Kalman Filter
摘要：
为了拍摄暗淡的系外行星和盘，配备日冕仪的天文台需要焦平面波前校正以恢复高对比度。最有效的校正方法是迭代地估计恒星电场，并用有源光学器件抑制它。该估计每次迭代需要来自科学相机的若干图像。为了最大限度地提高科学产出，需要进行快速波前校正并利用所有校正图像进行科学目标检测。系外行星和盘与它们的恒星是不相干的，因此需要一个非线性估计器来估计非相干强度和恒星电场。这种技术具有高度稳定性，只有在天基观测站和可能的地面望远镜上才能找到这种稳定性，这些望远镜具有极强的自适应光学。在本文中，我们实现了一个非线性估计，迭代扩展卡尔曼滤波器（IEKF），使快速波前校正和递归，近最佳估计的非相干光。在普林斯顿大学的高对比度成像实验室，我们证明，IEKF允许波前校正至少与卡尔曼滤波器一样快，并提供最准确的检测微弱的同伴。非线性IEKF形式主义允许我们追求其他策略，如参数估计，以提高波前校正。
英文摘要：
For imaging faint exoplanets and disks, a coronagraph-equipped observatory needs focal plane wavefront correction to recover high contrast. The most efficient correction methods iteratively estimate the stellar electric field and suppress it with active optics. The estimation requires several images from the science camera per iteration. To maximize the science yield, it is desirable both to have fast wavefront correction and to utilize all the correction images for science target detection. Exoplanets and disks are incoherent with their stars, so a nonlinear estimator is required to estimate both the incoherent intensity and the stellar electric field. Such techniques assume a high level of stability found only on space-based observatories and possibly ground-based telescopes with extreme adaptive optics. In this paper, we implement a nonlinear estimator, the iterated extended Kalman filter (IEKF), to enable fast wavefront correction and a recursive, nearly-optimal estimate of the incoherent light. In Princeton's High Contrast Imaging Laboratory we demonstrate that the IEKF allows wavefront correction at least as fast as with a Kalman filter and provides the most accurate detection of a faint companion. The nonlinear IEKF formalism allows us to pursue other strategies such as parameter estimation to improve wavefront correction.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
