中文摘要：

探索Blazar光变资料中的周期或准周期变化是1个有待深入研究的领域．由于光变资料的复杂性，目前寻找周期的算法还不够完善．文中以现代谱估计为基础，详细论述了自回归（Auto-regressive，AR）模型谱估计方法和最大熵谱估计的基本原理，分析了阶数选择对模型的重要影响，并把这些方法应用到类星体3C279和BLLac天体OJ287的光变周期分析中，得到它们的光变周期分别为7．14和11．76yr．通过验证，自回归模型谱估计方法由于其分辨率高，可以作为1种较好的分析周期的方法．最后指出在应用谱估计分析Blazar天体的光变周期注意事项．

英文摘要：

The periodicity analysis of Blazars＇ light curves needs improvement. It is believed that variability is one of the defining characteristics of Blazars. Due to the complication of Blazars＇ light curves, the present periodicity analysis methods are not perfect. Power spectrum analysis is available if the underlying variability process is stationary, which means that the process does not change with time in statistics. Classical spectral analysis methods are unsatisfactory to solve the problem. Based on modern spectral estimation, the principle of maximum entropy estimation and Auto-regressive（AR） spectral estimation are discussed. We mainly analyze the effect of the order selection on models. Therefore, the programme applying maximum entropy estimation and AR spectral estimation to deal with the period of quasar 3C 279 and BL Lac Object OJ 287 is presented. Their derived periods of 3C 279 and OJ 287 are 7.14 and 11.76 yr, respectively. Validated by the experiments, the AR model spectrum estimation method is a good way to compute the period of light variability. At the end of this paper, we point out what should be paid attention to in the applications.