Design of a Statistical Control Chart for Simultaneous Monitoring and Fault Isolation of Mean Vector and Covariance Matrix of Multivariate Multistage Processes

Pirhooshyaran, Mohammad | 2014

597 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 46654 (01)
  4. University: Sharif University Technology
  5. Department: Industrial Engineering
  6. Advisor(s): Akhavan Niaki, Taghi
  7. Abstract:
  8. In modern industries, multivariate multistage auto-correlated processes are widely used to ensure productivity and product quality. Interconnections between work stations bring a challenging task in detecting various shifts and identifying their root causes. In addition, simultaneous monitoring process mean and variability with single control chart methods has gained considerable attention throughout these years. In this article, a double-max multivariate exponentially weighted moving average (DM-MEWMA) chart is proposed based on two novel statistics to monitor the parameters of multivariate multistage auto-correlated processes jointly. Prior knowledge of variation propagation has been used in a way that the chart has both a fault identification power and a capability of working with sample size of one. Statistical test shows that the two statistics are independent of the process dimension. Monte Carlo Simulation indicates that DM-MEWMA chart has quite robust performance to detect changes and moreover, when the number of stages increases, it outperforms some existing alternative methods. In addition, fault identification comparison demonstrates that most of moderate mean and variability shifts can be isolated by the DM-MEWMA chart
  9. Keywords:
  10. Statistical Quality Control ; Exponentially Weighted Moveing Average (EWMA) ; Individual Observations ; Simultaneous Monitoring ; Fault Identification

 Digital Object List