A New Non-Parametric CUSUM Control Chart For One-Sided Monitoring

Control chart is a very good tool for statistical process control(SPC)to monitor the quality of the production process.Most of the traditional control charts used for production process monitoring assume that they have a full understanding of the process distribution.This assumption is unrealistic in many cases.In order to make up for the defect of parameter control chart,people began to study the control chart without specifying any process distribution.Nonparametric control chart does not need to specify any process distribution so it can better meet production requirements.However,most of the existing nonparametric control charts are difficult to distinguish different drift directions,which will affect the monitoring effect of asymmetric out of control process and target production process with known out of control direction.Therefore,this paper proposes a one-sided nonparametric CUSUM control chart based on Gaussian kernel function,which improves the efficiency of one-sided monitoring to meet the actual production needs.In the part of numerical simulation,this paper simulates the monitoring effect of the control chart when the observed values obey different distributions,and compares it with other control charts,and finds that it has better monitoring effect.The simulation results show that the proposed method has certain monitoring advantages.Finally,it is applied to an example to illustrate its applicability.