天文光变周期提取算法综述  

Review of Periodicity Searching Algorithms of Astronomical Light Curves

作　　者：安涛[1,2] 王俊义[3] 陆相龙[1,3] 劳保强 魏延恒 董典桥 陆扬[1] 伍筱聪 

机构地区：[1]中国科学院上海天文台,上海200030 [2]中国科学院射电天文重点开放实验室,南京210008 [3]桂林电子科技大学广西密码学组信息安全重点实验室,桂林541004

出　　处：《天文学进展》2016年第1期74-93,共20页Progress In Astronomy

基　　金：科技部政府间科技合作专项"SKA科学数据处理关键技术研究";科技部973项目"SKA建设准备阶段关键问题研究"(2013CB837900)

摘　　要：随着大量巡天项目的开展和完成,天文领域积累了大量的光变数据。这些数据中蕴含了许多有重要价值的物理参数,如光变周期,它对研究光变的物理机制、爆发预测和天体质量估算等有很大的帮助。但受到天体自身动力学过程、观测条件和仪器状态等客观因素的影响,光变数据往往是非均匀采样的,并且都不同程度地受到噪声的影响。因此,非常有必要研究基于非均匀采样信号的光变周期提取算法。综述了基于频域、时域和时-频域分析的三大类光变周期提取算法,主要包括Lomb-Scargle周期图法、Jurkevich算法和加权小波Z变换算法等,着重分析和比较了这些算法的性能,同时分析了它们的优缺点,最后总结与展望了光变周期算法。In astronomy, large amounts of light curve data have been accumulated as the development of increasing number of survey projects. The research of those light curve data is vital as they carry plenty of information about many important physical parameters. The study on the variability of light curve, for example, is very helpful for understanding the physical mechanism of light curves, forecasting outbursts, estimating celestial object mass and so on. Traditional variability study based on Fourier transform works well for time series data that are equally spaced in time, while that is not necessarily the case for light curve data since they are usually unevenly sampled due to various factors such as the dynamics of celestial objects, observational and instrumental condition and so on. Besides, light curve data are normally affected by noise to various extent. Therefore, algorithms based on uneven sampling ought to be explored and applied for the study of light curves. This paper summarizes the three types of numerical techniques for identifying periodicity of light curves in frequency domain, time domain and time-frequency domain, which mainly includes Lomb-Scargle periodogram, Jurkevich method, Weighted Wavelet Z Transformation, and so on. The performance of those algorithms is analyzed and the advantages and drawbacks of each of them are reviewed. A conclusion and discussion is provided in the end based on the analysis and comparison of the above algorithms.

关 键 词：非均匀采样时间序列 周期 频域 时域 时一频域 

分 类 号：P113[天文地球—天文学]