SOME EWMA SCHEMES for SIMULTANEOUS MONITORING of UNKNOWN MEAN and VARIANCE of a NORMALLY DISTRIBUTED PROCESS

• Activity Number: 670 - Advances in Statistical Process Control
• Type: Contributed
• Date/Time: Thursday, August 2, 2018 : 10:30 AM to 12:20 PM
• Sponsor: Quality and Productivity Section
• Author(s): Ridwan Sanusi* and AMITAVA MUKHERJEE and Min Xie
• Companies: City Univ of Hong Kong and INDIAN INSTITUTE OF MANAGEMENT and City University of Hong Kong
• Keywords: Distance-scheme; EWMA; Max-scheme; Normal-distribution; Reference-sample; Simultaneous-monitoring

Abstract

In the production line of a manufacturing industry, one should monitor both the average and spread of an underlying process to ensure improved item quality. Several recent articles established that monitoring several quality characteristics using a single plotting statistic has some added advantage over using two or more separate process monitoring schemes. The normal distribution is widely used in the context of quality monitoring. Noting that, we introduce 3 EWMA schemes based on single plotting statistics for simultaneously monitoring the mean and variance of a normally distributed process. Further, we assume that the process parameters are unknown a-priori, which is most common in practice. We consider a reference sample from a suitable in-control population and use them to estimate unknown process parameters. We take into account the effect of estimation in constructing the EWMA schemes. We propose a max-type EWMA scheme and 2 distance-type EWMA schemes and compare them with some relevant existing charts. We highlight the design and implementation strategies of the proposed schemes using real data sets from a forged automobile engine piston rings and from a timber industry.