The Application Of Multivariate Process Control Charts

In order to meet the diversified requirements of products in production and life,the formation of a product needs multiple processes.The multivariate control chart is applied to monitor the key processes.Once abnormal is discovered,measures should be taken in time to reduce losses,improve efficiency and ensure quality.In particular,it is often necessary to monitor multiple elements at the same time in reality,so it is necessary to study and develop multiple control charts.For this reason,two kinds of multiple control charts are used in this paper: Multivariate Exponentially Weighted Moving(MEWMA)control chart,and A Spatial-Rank Multivariate EWMA(SREWMA)control chart.The application of multivariate process control charts is mainly studied from three aspects:First,there are a variety of cardiopulmonary function self-test method: floor experiment,exercise testing,or measuring pulse and blood pressure,etc.,then use a variety of test systems or monitoring models to determine whether they are in a healthy state,but none of these methods can be used to alert people before the onset of cardiopulmonary function,and the indicators are complex.Considering the diversity of cardiopulmonary function parameters and change slowly in pathology,,using the MEWMA control chart which is sensitive to small and medium drift to monitor the cardiopulmonary function.The control chart can accumulative amplifying tiny changes of cardiopulmonary function,give an alarm before cardiopulmonary disease,timely understand whether it is in a state of health,and medical check in time,avoid missed the best treatment time,have a very good warning effect.In addition,it can reduce the frequency and cost of physical examination and improve the utilization rate of hospital resources.Second,application of MEWMA control chart can well monitor the process of small and medium-sized drift,but it requires monitoring data obey the multivariate normal distribution.In fact,it is difficult to meet the requirement.Therefore,the SREWMA control chart which does not need to meet the specific distribution is introduced.Money is an important medium of material exchange in people’s life,and its authenticity is significant.so it is necessary to monitor banknote quality and reduce abnormal banknotes in production.In this paper,SREWMA control chart is used to monitor the quality of banknotes,when abnormal factors are in trend but there are no unqualified banknotes,timely alarm and measures are taken to reduce the waste ofmanpower and material resources.Third,during the application of SREWMA control chart,we find that with the increase of data,the number of computation space rank is increasing,and we need to keep historical data,the memory needs more and more.This is not applicable to the big data of large production line.In this paper,the concept of storage space is added to SREWMA control chart,that is,after the observed value reaches a certain amount,the number of spatial ranks is fixed,and its computation and operation memory are reduced.The improved SREWMA control chart also satisfies the monitoring at the beginning of the process,and has good robustness for various distributions and drifts,and greatly improves its computation speed.Finally,the two control charts are compared and analyzed by the example of the quality control of banknote,and the actual value is verified.