Detection and identification of mean shifts in multivariate autocorrelated processes: A comparative study

Abstract

Complex processes with autocorrelated multivariate quality characteristics often exist in business/industry. However, limited researches have been done in detecting and identifying process mean shift in multivariate autocorrelated processes. In this thesis, a neural-network based control scheme is proposed to simultaneously detect and identify mean shifts in multivariate autocorrelated processes. The proposed control scheme utilizes the effective Extended Delta-Bar-Delta learning rule and is trained with the powerful Back-Propagation algorithm. To illustrate the power of the proposed control scheme, its Average Run Length (ARL) performance is evaluated against three statistical control charts, namely, the Hotelling T-square chart, the MEWMA chart, and the Z chart, in multivariate autocorrelated processes. It is shown that the NN-based control scheme can detect and identify mean shifts effectively and efficiently in multivariate autocorrelated processes.