Abstract: Traditional ARMA residual control charts are often sensitive to outliers, which easily leads to failures in monitoring. In order to solve this problem, this article uses the idea of robust statistics to revise the traditional ARMA residual control charts, and constructs a robust ARMA residual control chart algorithm to overcome the influence of outliers on the model. From the simulation and empirical results, it is known that when there are no outliers in the data, the monitoring results obtained by the traditional and robust ARMA residual control charts are basically the same; when there are outliers in the data, compared to the traditional ARMA residual control charts the robust ARMA residual control charts can more effectively resist the influence of outliers, and have better anti-interference and high tolerance.

Key words: Robust ARMA model , Residual control chart, Outliers