|
Activity Number:
|
487
|
|
Type:
|
Invited
|
|
Date/Time:
|
Thursday, August 2, 2007 : 8:30 AM to 10:20 PM
|
|
Sponsor:
|
Section on Quality and Productivity
|
| Abstract - #307657 |
|
Title:
|
Monitoring Correlation Within Nonlinear Profiles Using Mixed Models
|
|
Author(s):
|
Willis Jensen*+ and Jeffrey B. Birch
|
|
Companies:
|
W.L. Gore & Associates, Inc. and Virginia Polytechnic Institute and State University
|
|
Address:
|
3750 West Kiltie Lane, Flagstaff, AZ, 86003-2400,
|
|
Keywords:
|
Multivariate Statistical Process Control ; Nonlinear Model ; Phase I ; Blups ; T-squared Statistic
|
|
Abstract:
|
Profile monitoring is a technique in quality control best used where process data follow a profile at each time period. Previous work on monitoring of nonlinear profiles has assumed that the measurements within a profile are uncorrelated. To relax this restriction we propose the use of a nonlinear mixed model (NLMM) to monitor the nonlinear profiles to account for the correlation structure. We evaluate the effectiveness of fitting separate nonlinear regression models to each profile in Phase I control chart applications for data with uncorrelated errors and no random effects. We make the same evaluation for data with random effects. Our proposed approach uses the separate nonlinear regression model fits to obtain the NLMM fit. This NLMM approach results in charts with good abilities to detect changes in Phase I data and has a simple to calculate control limit.
|
