Abstract: This paper introduces the design and implementation of an automatic processing software system for LAMOST celestial spectra. Some key algorithms in the system such as covering algorithm, wavelet transform, mean shift, etc., are also presented. The experiment results over the datasets of sloan digital sky survey (SDSS) show that this system has good characters in processing speed, classifying accuracy and redshift measuring precision. And the implemented system is workable for the large survey projects.

Key words: LAMOST, System design, Covering algorithm, Wavelet transform, Mean shift

摘要： 介绍了LAMOST光谱数据自动处理软件系统的设计及实现。阐述了系统实现中所使用的核心算法：基于覆盖算法的光谱分类方法，基于小波变换的晚型恒星识别算法和基于均值漂移的红移求取算法。在美国SDSS的天体数据库上测试表明，该软件系统具有较快的处理速度，并且能够获得较高的分类正确率和红移计算精度，可以满足大型巡天计划的实际需求。

关键词: 大天区面积多目标光纤光谱天文望远境, 系统设计, 覆盖算法, 小波变换, 均值漂移