The article considers the possibility of maintaining in statistical controllable state the technological process with the use of control charts of accumulated sums. The principles of construction of these charts, features of the process disorder detection are shown. The feature of the control charts of accumulated sums is that, the decision regarding the statistical controllability of the process is made taking into account the preliminary information. The dots mapped do not correspond to individual measurement results or statistics, they display information about all the results obtained, from first to last, inclusive. This feature allows you to respond more quickly to signs of process disorder than when working with Shewhart control charts, where sample statistics are used for analysis, which are independent control results.
The characteristics relation of the charts with the establishment of errors of the 1st and 2nd kind when interpreting them using the hypothesis testing apparatus is considered. The article presents the theoretical substantiation and approaches to the application of sequential LQ (likelyhood quotient) criterion in process monitoring, describes the main parameters that influence decision making. A graphical interpretation of the application of the LQ criterion is given to improve the perception of information.
The method of graphical estimation of the process disorder with the use of V-mask, which is associated with the reversed LQ criterion, is also considered, if a disorder of the technological process is detected and it is necessary to set the time when it happened. At the same time, the right approach to constructing a V-mask template is very important, since in addition to the «opening» angle of the mask, the scaling factor must be taken into account. This coefficient is the ratio of the scale of the ordinate to the abscissa axis, and its failure to apply will lead to incorrect interpretation of the control results. A specific numerical example with an appropriate graphical representation of the results is considered.
This work is licensed under a Creative Commons Attribution 4.0 International License.