针对在MAP中, 测控设备高频率地采集数据导致数据呈现时序依赖性, 难以满足传统控制图要求的数据间相互独立的问题, 提出根据时间序列原理用受控状态下观测数据估计模拟出自回归模型AR(p),求出服从独立同分布的观测值残差, 建立控制图。为进一步说明残差控制图的有效性, 通过真实数据求出残差控制图并与休哈特控制图进行比较, 结果表明, 残差控制图的虚假警报明显减少。

In the Measurement Assurance Program (MAP), due to the high-frequency data acquisition of the measurement and control equipment, the data may present dependency on timing sequence, and can hardly meet the data independence requirements of the traditional control chart.It is proposed to simulate the auto-regression model AR (p) with the observation data in the constrained state according to the timing sequence principle, the residual of the observation value is calculated, which is independently and identically distributed, and the control chart is set up. To further illustrate the validity of the residual control chart, real data is used to obtain the residual control chart and then it is compared with the Shewhart control chart.The results show that the occurrence of the false alarm of the residual control chart is significantly reduced.