基于机器学习的快速射电暴搜寻方法综述  

A Review of Fast Radio Burst Search Methods Based on Machine Learning

作　　者：刘艳玲[1,2,3,4] 陈卯蒸 袁建平[1,2,3] Liu Yanling;Chen Maozheng;Yuan Jianping(Xinjiang Astronomical Observatory,Chinese Academy of Sciences,Urumqi 830011,China;University of Chinese Academy of Sciences,Beijing 100049,China;Key Laboratory of Radio Astronomy,Chinese Academy of Sciences,Nanjing 210033,China;Xinjiang Key Laboratory of Microwave Technology,Urumqi 830011,China)

机构地区：[1]中国科学院新疆天文台,新疆乌鲁木齐830011 [2]中国科学院大学,北京100049 [3]中国科学院射电天文重点实验室,江苏南京210033 [4]新疆微波技术重点实验室,新疆乌鲁木齐830011

出　　处：《天文研究与技术》2022年第5期509-517,共9页Astronomical Research & Technology

基　　金：国家自然科学基金(11903071)资助。

摘　　要：快速射电暴(Fast Radio Burst,FRB)是目前射电天文领域的主要热点前沿,相关研究被《自然》(Nature)杂志评选为2020年十大科学发现之一。快速射电暴爆发时间极短且鲜少重复的特点,使其被观测捕捉到的概率极低。由人工从海量的天文观测数据中识别快速射电暴是件耗时费力的工作。机器学习技术的蓬勃发展为实时搜寻与多频段联合跟踪观测快速射电暴带来了可能。从传统机器学习方法和深度学习方法两方面,对该研究已有的成果进行分析与总结,并探讨了目前基于机器学习的快速射电暴搜寻技术存在的问题和面临的挑战,分析了其未来发展趋势。在不久的将来,深度学习技术将更广泛地应用于快速射电暴搜寻,并成为实现高效搜寻快速射电暴的利器。Fast Radio Bursts(FRBs) are a hot topic in the field of astronomy at present. Its related research was also selected by the journal Nature as one of the top 10 scientific discoveries of 2020. The characteristics that FRBs are millisecond-duration and rarely repeated make them hard to be captured. Identifying FRBs from massive astronomical observation data by human review is a time-consuming and laborious task. With the rapid development of machine learning technology, it is possible to carry out a real-time search and multi-frequency tracking for FRB events. This paper analyzes and summarizes the existing representative results from two aspects: traditional machine learning method and deep learning method. Finally, the existing problems and challenges of FRB search technology based on machine learning are discussed, and future development trend is also analyzed. In the near future, deep learning technology will be more widely used and become a powerful tool to search for FRBs efficiently.

关 键 词：快速射电暴 机器学习 搜寻方法 深度学习 射电天文 

分 类 号：TP274[自动化与计算机技术—检测技术与自动化装置]