Abstract
In this paper, using multivariate engineering process control (MEPC) alone is compared with using MEPC coupled with multivariate statistical process control (MSPC). MEPC and MSPC are two strategies for quality improvement that have developed independently. MEPC is to minimize variability by adjusting some manipulative process variables to keep the process output on the target vector. MSPC is to reduce variability by monitoring and eliminating assignable causes of variation. The simulation is used to evaluate the average run lengths and the averages of the performance measure. The simulation results show that use of both MEPC and MSPC can always outperform the use of either alone. Especially in detecting small sustained shifts of the mean vector, combining MEPC with the multivariate generally weighted moving average (MEPC/MGWMA) chart is more sensitive than MEPC/multivariate exponentially weighted moving average (MEPC/MEWMA) chart and MEPC/Hotelling multivariate chart. An example of the application is also given.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call