We designed and built a laboratory model of a sparse space telescope. The purpose of the model is to test various algorithmic, optical, and mechanical issues with such a telescope: how to achieve alignment after deployment that yields diffraction limited optical performance. Our model uses four non-redundant sparse sectors. Each segment has three degrees of freedom, with supplementary coarse focusing, so the search volume for a perfect image is quite large, ~1051. We tried stochastic searches, optimizing the point spread to its minimum, and are now testing different directed search algorithms, of volume order 1019 or less. This will cut the search time from one day to about one hour, with much less chance of actuator failure.
|