发表机构
Université Paris-Sud XI; CNES — Centre National d’Études Spatiales; Onera — The French Aerospace Lab（巴黎第十一大学; 法国国家空间研究中心; 法国航空航天实验室）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
该研究表征了PERSÉE消零干涉仪测试台，采用LQG控制器抑制多频振动，实现创纪录消零率，验证了空间系外行星观测的可行性。
AI 中文摘要
观测系外行星受到高恒星对比度和小角间距的阻碍，消零干涉术通过重组两个光瞳使恒星发生相消干涉、行星发生相长干涉来解决这一问题。然而，要在外部干扰下维持深度消零，需要亚纳米级的光程差（OPD）稳定性和百分之一艾里斑级的指向控制。为验证该技术用于空间任务的可行性，由CNES领导的联盟开发了PERSEE实验室演示测试台，并集成于默东天文台。PERSEE模拟完整的空间任务配置，目标是在模拟干扰下，数小时内实现1e-4的稳定消光率，且变化量为1e-5。本研究聚焦于该测试台关键组件和共相控制回路的多级集成、校准与特性表征。通过实施经初步干扰测量优化的线性二次高斯（LQG）控制器，系统抑制了1-100Hz、振幅数十纳米的多频振动，将残余OPD降至0.3nm均方根，指向误差降至艾里斑的0.4%。这些稳定控制在1.65-2.45μm光谱带内，数小时内实现了8.8e-6的创纪录消零率，稳定性达9e-7，超出初始规格一个数量级。将这些结果外推至空间任务，表明配备40cm望远镜和100Hz控制回路时，可对亮于9等星的恒星开展系外行星观测。
英文摘要
Observing exoplanets is hindered by high stellar contrast and tiny angular separation. Nulling interferometry addresses this by recombining two pupils for destructive interference on the star and constructive interference on the planet. However, maintaining deep nulls despite external disturbances requires sub-nanometer optical path difference (OPD) stability and sub-hundredth Airy disk pointing control. To validate this technology for a space mission, the PERSEE laboratory demonstrator bench was developed by a CNES-led consortium and integrated at the Meudon Observatory. Simulating a complete space mission setup, PERSEE targeted a stable star extinction rate of 1e-4 with variations of 1e-5 over several hours under simulated disturbances. The thesis focused on the multi-stage integration, calibration, and characterization of the bench's critical components and cophasing control loops. By implementing a Linear Quadratic Gaussian (LQG) controller optimized through preliminary disturbance measurements, the system mitigated multi-frequency vibrations (1-100 Hz, tens of nanometers amplitude), reducing residual OPD to 0.3 nm RMS and tip-tilt errors to 0.4 percent of the Airy disk. These stabilization controls achieved a record null rate of 8.8e-6 with 9e-7 stability over several hours in the 1.65-2.45um spectral band, surpassing initial specifications by an order of magnitude. Extrapolating these results to space missions indicates that with 40 cm telescopes and 100 Hz control loops, exoplanet observations are feasible for stars brighter than 9th magnitude.
CommentsPhD thesis defended in 2012, originally written in French. Translated to English with the help of Google/Gemini. With the revival of nulling interferometry projects, I felt it was timely to get an English version out
