目的　给出一种有效的处理含缺失值时间序列的方法 ,完成缺失值的内插及ARMA模型的参数估计。方法　用状态空间的Markov表达描述时间序列 ,进而采用Kalman滤波技术。结果　实例分析表明 ,不仅可以完成缺失值的有效内插 ,模型拟合效果及预测结果也甚为满意。结论　用基于状态空间表达的Kalman滤波技术 ,可以实现平稳可逆时间序列中缺失值的内插及ARMA模型拟合 Objective To give an effective method to process time series with missing values, complete the interpolation of missing values ​​and estimate the parameters of ARMA model. Methods The Markov representation in the state space is used to describe the time series, and then the Kalman filtering technique is used. Results The example analysis shows that not only the effective interpolation of missing values ​​can be completed, but also the model fitting effect and the prediction result are very satisfactory. Conclusion The Kalman filtering technique based on state space expression can realize the interpolation of missing values ​​in stationary reversible time series and the fitting of ARMA model.