Design And Diagnosis Of Nonparametric Control Chart Based On Cucconi Test

Improving product quality is an important guarantee for enterprises to improve market competitiveness.Statistical quality control chart is widely used in the production site and is an important means of quality control in the production process.In the past five years,research on non-parametric process monitoring has grown tremendously.Because nonparametric control charts do not require strict assumptions,they can monitor various abnormal or complex process changes very effectively.The traditional normal process monitoring scheme uses two separate control charts,one is used to monitor the change of position parameters and the other is used to monitor the change of scale parameters.In this paper,Cuccconi statistics is used to decompose them into three different forms.Combined with the circular grid control chart,the change of position parameters and scale parameters are monitored simultaneously,and the out of control(OOC)signal is visualized,This paper also compares the performance of three decomposition cases,and finds that different decomposition methods and different data types have different effects on detecting the source of data parameter changes,Combined with the data type,the correct decomposition method is selected to effectively detect the parameter changes of the data.Through Monte-Carlo simulation,we compared our proposed scheme with several existing control schemes,and used the average run length to reflect the performance of the control chart.Through comparison and simulation,we found that small changes were found in the position and scale parameters.When the time,the monitoring effect of our proposed plan will be better.Finally,an example is given to prove the practical application of circular grid control chart.