COMPARING CONTROL CHARTS FOR GAUSSIAN MEAN VECTORS WITH MEWMA AND SLIDING WINDOW SCHEMES

In this work we show that for normal distributions the Hotelling ́s T and the multivariate exponentially weighted moving average (MEWMA) distances are directly related to the Bhattacharyya distance. This relationship pr ovides important information concerning on the misclassification error probability as an up per bound on it. In fact, this useful information indicates the overlap degree between th e inand out-of-control processes. Therefore, the first purpose of this simulation stu dy is to monitor the mean vector of a bivariate Gaussian process by means of an informati ve control chart based on probability bounds. Additionally, a comparison study is carried to measure the effects of estimating the actual mean vector by the MEWMA scheme and sliding w dow schemes, which are chosen to have uniform, linear and exponentially distributed weights. Results demonstrated that the confidence MEWMA control chart is easier to calibra te nd shows less inertia for big shifts in the mean vector than the sliding window approach.

[1]  H. Hotelling.  Multivariate Quality Control-illustrated by the air testing of sample bombsights , 1947 .

[2]  C. Rao.  The Problem of Classification and Distance Between Two Populations , 1947, Nature.

[3]  Calyampudi R. Rao.  On the distance between two populations , 1949 .

[4]  R. Crosier.  Multivariate generalizations of cumulative sum quality-control schemes , 1988 .

[5]  Rolf Isermann,et al.  On-line Identification in Closed Loop , 1991 .

[6]  Charles W. Champ,et al.  A multivariate exponentially weighted moving average control chart , 1992 .

[7]  N F Hubele,et al.  X¯ CONTROL CHART PATTERN IDENTIFICATION THROUGH EFFICIENT OFF-LINE NEURAL NETWORK TRAINING , 1993 .

[8]  Norma Faris Hubele,et al.  Back-propagation pattern recognizers for X¯ control charts: methodology and performance , 1993 .

[9]  Yeou-Ren Shiue,et al.  On-line identification of control chart patterns using self-organizing approaches , 2005 .

[10]  Ruey-Shiang Guh,et al.  An effective application of decision tree learning for on-line detection of mean shifts in multivariate control charts , 2008, Comput. Ind. Eng..

[11]  S. T. A. Niaki,et al.  A new statistical process control method to monitor and diagnose bivariate normal mean vectors and covariance matrices simultaneously , 2009 .

[12]  Ahmed Ghorbel,et al.  A survey of control-chart pattern-recognition literature (1991-2010) based on a new conceptual classification scheme , 2012, Comput. Ind. Eng..

[13]  Alireza Faraz,et al.  Multiobjective Genetic Algorithm Approach to the Economic Statistical Design of Control Charts with an Application to X¯ bar and S2 Charts , 2013, Qual. Reliab. Eng. Int..