An Adaptive T² Chart for Monitoring Dynamic Systems

Abstract

Although most statistical process control techniques are designed to detect constant process shifts, time-varying shift patterns are more frequently encountered in industrial practice. In this paper, we propose an adaptive T² scheme for monitoring processes with dynamic shifts. The new scheme preserves the optimality of directionally variant charts by updating a reference mean-shift vector recursively and can be easily adjusted to obtain high sensitivity across desired shift ranges. Simulation studies show that the adaptive T² chart with EWMA forecasting of mean shifts outperforms most conventional charts in a dynamic environment and is also robust to parameter-estimation uncertainties.

• Adaptive Chart
• Directionally Variant Chart
• Exponentially Weighted Moving Average
• Hotelling's T² Statistic
• Process Dynamics

Additional information

Notes on contributors

Kaibo Wang

Dr. Wang is an Assistant Professor in the Department of Industrial Engineering. His email address is [email protected].

Fugee Tsung

Dr. Tsung is an Associate Professor in the Department of Industrial Engineering and Logistics Management. He is a Senior Member of ASQ. His email address is [email protected].