Exclusive Content & Downloads from ASQ
An Adaptive T-Squared Chart for Monitoring Dynamic Systems
Summary: [This abstract is based on the authors' abstract.] An adaptive T² scheme is proposed for monitoring processes with dynamic shifts that are frequently encountered in an industrial setting. The approach preserves the optimality of directional variant charts by updating a reference mean-shift vector recursively, and it can be adjusted to obtain high sensitivity across shift ranges. Simulation studies show that the proposed chart with exponentially weighted moving average forecasting of mean shifts outperforms most conventional charts in a dynamic environment and is robust to parameter-estimation uncertainties.
- Topics: Software Quality, Software and Technology (for statistics, measurement, analysis)
- Keywords: Exponentially weighted moving average control charts (EWMA), Hotelling's T2 statistic, Process dynamics, Shift point
- Author: Wang, Kaibo; Tsung, Fugee
- Journal: Journal of Quality Technology