时间序列的应用在科学领域内愈加广泛。本篇论文以时间序列模型为基础，包括AR模型、MA模型和ARMA模型，以比较其各种参数识别方法的区别和联系为目的。首先对相关概念和定义以及模型形式作了介绍，然后通过计算各模型的自相关函数的算法引出了参数矩估计的方法，而后重点分析了参数估计的极大似然法和最小二乘法，并从系统控制理论方面阐述了最小二乘的递推公式和加权递推公式以及广义最小二乘的方法，从而填补了非正态情况下参数估计的空白，完整了整个参数辨识的框架。最后对以上这些参数估计方法作了对比，采用应用具体实例的途径，揭示了各个方法的区别、优劣和有效性等

As the Time Series Analysis developed more and more widely in the scientific application areas, the method of mathematical modeling on it is of great importance. This paper is focused on one of the key step of modeling, the identification and estimation of the parameters.Based on some related conception and definition, the initial estimation is educed by the algorithm of autocorrelation function, which is one of the basic and regular estimation method. Later I introduce a more efficient and commonly used approach, the maximum likelihood estimation and least square estimation,followed by the recursion least square function and the generalized one.The latter one has taken up the blank space of the Non-Normal distribution series. And finally come to the central part of the paper,the discrimination, which has give a general comparison of the superiority and inferiority of the various methods, in the ways of giving example.