对休哈特控制图中平均值控制界限的商讨

Discussion on Mean Value Control Boundary of Shewhart Control Charts

采用休哈特控制图的方法可以对测量过程是否处于统计控制状态进行控制。但是，对于平均值控制界限的确定则认为不妥，根本原因是，平均值控制界限依据重复性条件下所得数据的标准偏差来设置，用它来控制复现性条件下子组的平均值的波动性，由于条件不同，因此，前者不能来控制后者。依据计量保证方案（MAP）提出的控制界限的确定，是用在复现性条件下各子组的平均值的组间标准差来设置控制界限，也就是用组内平均值的分散性来控制组内平均值。通过对二等电阻、直流电压和电能表标准装置同一组试验数据进行上述两种控制图的制作比较，以及对确定控制界限的理论分析，证明了休哈特控制图平均值控制界限的确定不妥。

The Shewhart control chart can be used to control the measuring process whether it is under the state of statistical control. But it is considered that the confirmation for the control upper and lower limits of the mean value is improper. For the reason that the upper and lower control limits is established from standard deviation of measuring data under repeated conditions. And then, using it to control the mean value＇s variations character of sub -groups, in fact, the former cannot control subgroup variation due to the condition is different. Other way is described to confirm the upper and lower control limits which is based on measurement assurance programs （MAP）. This method is to use sub-groups＂ standard deviation obtained from repeated conditions to setting the control limits, i.e. the dispersion of mean value within the subgroup to control the mean value. It has proved that the determination of the mean value controlling boundary via Shewhart control chart is improper by means of comparing the Shewhart with MAP method during analyzing the same experimental data of the DC resistance, DC voltage and electricity meter test equipments.