SIMULATION STUDIES PERFORMANCE OF EWMA-MAX MCHART BASED ON SYNTHETIC DATA

Abstrak

Quality control has an important role in the manufacturing process. One of the statistical tools used in quality control is Statistical Process Control (SPC). The SPC product is a control chart. A control chart is a graphical tool used to determine if a process is under statistical quality control, helping to identify issues and drive quality improvements. Control charts are usually used to control variables or attribute data quality. Commonly used variable data is data with mean and variability characteristics. Various types of control charts are control charts for mean, control charts for variability, and simultaneous control charts designed to control mean and variability simultaneously. In real-field practice, manufacturing requires multivariate process control because many variables must be controlled. This research proposes a multivariate simultaneous control chart, the Exponentially Weighted Moving Average Max Multivariate (EWMA Max-Mchart). This control chart can handle multivariate process control simultaneously, both process mean and process variability. This research tests the performance of control charts with a simulation study using synthetic data with several process mean conditions and a covariance matrix. As a comparison, the development of the previous Max-M control chart was also tested. Based on the synthetic data generated, a performance comparison was made by looking at the suitability of in-control and out-of-control. The comparison results show that the EWMA Max-Mchart has better quality control performance if there is a shift than the Max-Mchart.