Research On The Key Technology Of Fiber Position Measurement For The Spectral Survey Telescope

Large-scale spectroscopic survey telescopes can directly observe and obtain rich spectral information on celestial objects,playing an important role in astronomical and astrophysical research.The Large Sky Area Multi-Object Fiber Spectroscopic Telescope(LAMOST)is a representative large-field-of-view,large-aperture,and multiobject optical fiber spectroscopic telescope,which has been operating for 12 years.Its massive spectrum data ouput has greatly promoted the leapfrog development of astronomy in China.Initially,LAMOST used open-loop positioning to locate optical fibers.To further improve the positioning accuracy and operational reliability of the optical fiber positioning unit,the telescope is currently undergoing an upgrade of the optical fiber positioning detection system.To meet the needs of constructing the LAMOST optical fiber positioning detection system,this thesis proposes a new LAMOST optical fiber position detection benchmark framework.The designed benchmark fiber serves as a common reference point,and the relationship between object space and image space is calculated to ensure the accuracy of the framework.Different from the conventional use of reference points in photogrammetry,the proposed benchmark fiber can achieve micro-level detection accuracy under ultra-long distance(20 meters)conditions.Currently,the key technical issue in this field is how to achieve higher accuracy in fiber positioning.Compared with the latest sky survey project,DESI,the focal plane size of LAMOST is larger and the camera is farther away from the optical fibers.The diameter of LAMOST’s focal plane is 1.75 meters,and the distance between the camera and the optical fibers is 20 meters.In contrast,the diameter of DESI’s focal plane is 0.9 meters,and the distance between the camera and the optical fibers is 12.25 meters.Therefore,achieving high-precision optical fiber detection in LAMOST is a more challenging task.We have studied the design of fiducial fibers’ structure,quantity,and distribution,as well as the detection algorithm for fiducial fibers,to improve the accuracy of fiber positioning.The LAMOST closed-loop control system upgrade project requires that the detection accuracy of optical fibers reaches 20 microns and the positioning accuracy of optical fibers reaches 40 microns.To ensure that the optical fiber detection and positioning accuracy of LAMOST meet the needs of telescope sky survey observations,this thesis primarily conducted the following research work:(1)A scheme is proposed that uses fiducial fibers to correct visual detection errors and improve positioning accuracy,and a fiducial fiber structure that meets accuracy requirements and is easy to implement and verify is designed.On the one hand,the fiducial fiber can provide a reference for multiple shots of images in fiber positioning to reduce system errors during each shot,and on the other hand,it can calculate the object-image transformation relationship of camera-shot images based on the fiducial fiber to complete the conversion of working fiber coordinates from pixel coordinates to focal plane coordinates.To improve the similarity between the fiducial fiber and the working fiber and avoid introducing unnecessary fiber detection errors,the fiducial fiber is designed with the same core diameter and a similar illumination method as the existing working fiber.The position of the fiducial fiber unit on the focal plane can be accurately measured by a laser tracker,and the measured results have reached the expected accuracy requirements.(2)A fiducial fiber distribution algorithm based on the Voronoi diagram area standard deviation is proposed.As a reference benchmark,the fiducial fibers need to cover the entire surface being measured,and their quantity and distribution seriously affect the accuracy of the fiber position detection system.Since the fiducial fibers need to occupy the installation position of the working fibers,their quantity will be limited.To obtain a uniform distribution under limited quantity,this thesis uses the characteristics of Voronoi diagrams to quantitatively analyze and optimize the position of fiducial fibers,showing better performance compared with existing algorithms.This thesis also combines particle swarm algorithms for the iterative calculation to obtain the final distribution result of fiducial fibers.The distribution algorithm proposed in this thesis can use relatively few fiducial fibers for fiber detection,laying a hardware foundation for establishing a high-precision reference framework for fiber position detection systems.(3)An algorithm for identifying,encoding,and decoding fiducial fibers is proposed.Distinguishing between the fiducial and working fibers can be difficult because both appear as bright spots in the image.To address this issue,we present an algorithm that combines concentric circle traversal and K-means clustering,providing unique codes for each fiducial fiber for encoding and quick recognition.A decoding algorithm is also designed to determine the position of each fiducial fiber in the focal plane coordinate system according to its code.The proposed algorithm achieves high accuracy and speed,meeting the demands of rapid identification of fiducial fibers.(4)A closed-loop fiber positioning detection system based on the fiducial fiber framework is designed and built for the LAMOST fiber positioning position detection upgrade and retrofit project,and a fast-working fiber matching algorithm is proposed.This system simplifies complex steps such as looking up pulse calculations for unit motors in the original open-loop positioning method and setting up cameras in front of the focal plane to calibrate all units during a large amount of maintenance time in summer every year.It can detect and correct unit positioning errors in real-time through cameras,improve positioning accuracy and observation efficiency,and monitor and screen faulty units.This thesis conducts research on high-precision fiber position detection technology in closed-loop fiber positioning technology,one of the core technologies of spectroscopic survey telescopes,and proposes and implements a closed-loop fiber positioning scheme based on fiducial fibers in the LAMOST fiber positioning system.This thesis focuses on analyzing the design,quantity,and distribution of fiducial fibers and the implementation of closed-loop positioning systems,and verifies the feasibility of the scheme through LAMOST field experiments.The relevant research results are of great significance for ensuring fiber positioning accuracy,improving spectral acquisition efficiency,and positioning 10,000 fibers,high-density,miniaturized fiber positioning systems in the LAMOST Phase Ⅱ project that has already begun preliminary research and project approval work.