Chen Shanyong, Wu Chuanchao, Zhai Dede, Liu Junfeng
Appl Opt. 2018 Nov 1;57(31):9397-9405. doi: 10.1364/AO.57.009397.
The membrane diffractive telescope is an alternative to future 10-meter class space telescopes due to its possible lower cost of manufacturing and launching. To actively control the misalignment, the edge sensor network is utilized to measure the relative in-plane motions as well as out-of-plane motions of neighboring segments. The measurement is vulnerable to noise and malfunction of sensors. We first model the segment kinematics, which is used to simulate the propagation of sensor error to the segment configuration error. Monte Carlo simulation results show that the sensor error is considerably accumulated in the configuration error, which follows approximately normal distribution. On the other hand, the ring-arranged segments imply the form closure with redundancy of edge sensors. It is employed to judge whether there are abnormal sensor readings and then to identify the malfunctioning sensors. A better estimation can further be obtained as correction to the abnormal readings. Simulation results show that without sensor error, the algorithm in most cases can identify two malfunctioning in-plane sensors and two malfunctioning out-of-plane sensors. While considering the sensor error, only one malfunctioning in-plane sensor and one malfunctioning out-of-plane sensor can be identified with a sufficiently big ratio of abnormal reading to the normal sensor error.
膜衍射望远镜因其可能较低的制造和发射成本，成为未来10米级空间望远镜的一种替代方案。为了主动控制对准误差，利用边缘传感器网络来测量相邻节段的相对平面内运动以及平面外运动。该测量容易受到噪声和传感器故障的影响。我们首先对节段运动学进行建模，用于模拟传感器误差向节段配置误差的传播。蒙特卡罗模拟结果表明，传感器误差在配置误差中会大量累积，配置误差近似服从正态分布。另一方面，环形排列的节段意味着边缘传感器具有冗余的形式封闭。它用于判断是否存在异常的传感器读数，进而识别出故障传感器。作为对异常读数的校正，可以进一步获得更好的估计值。模拟结果表明，在没有传感器误差的情况下，该算法在大多数情况下能够识别出两个故障平面内传感器和两个故障平面外传感器。在考虑传感器误差的情况下，当异常读数与正常传感器误差的比值足够大时，只能识别出一个故障平面内传感器和一个故障平面外传感器。
