LAMOST Fiber Positioning Unit Detection Based on Deep Learning

Abstract The double revolving fiber positioning unit (FPU) is one of the key technologies Large Sky Area Multi-Object Fiber Spectroscope Telescope (LAMOST). accuracy computer controlled FPU depends on robot as well initial parameters FPU. These may deteriorate with time when running in non-supervision mode, which would lead to bad position and further efficiency degradation subsequent surveys. In this paper, we present an algorithm based deep learning detect FPU’s angle using front illuminated image LAMOST focal plane. Preliminary test results show that detection better than 2.°5, good enough distinguish those obvious FPUs. Our are verified by direct measurement from back correlation analysis spectral flux survey data.