ADAPTIVE EXPONENTIALLY WEIGHTED MOVING AVERAGE WITH MEASUREMENT ERROR (COVARIATE) WITH AUXILIARY INFORMATION MAXIMUM FOR CEMENT QUALITY CONTROL

Keywords: Adaptive, Auxiliary Variable, Control Chart, Covariate, EWMA, Measurement Error

Abstract

The Shewhart control chartexhibits limitations in detecting small process shifts and monitors the mean and variance separately. To address these shortcomings, this study introduces the Adaptive EWMA with Measurement Error (Covariate Method) and Auxiliary Information Max (AEWMA ME C AI Max) control chart. This novel approach integrates memory-based monitoring, joint mean-variance detection, measurement error correction through the covariate method, utilization of auxiliary variables, and adaptive adjustment mechanisms to enhance sensitivity across various shift magnitudes. The AEWMA ME C AI Max chart was applied to cement production data from PT XYZ, using Blaine fineness as an auxiliary variable for monitoring compressive strength. Comparative analysis demonstrates that the adaptive chart consistently produces control statistics closer to the upper control limit compared to the non-adaptive Max-EWMA ME C AI chart, validating its superior sensitivity in shiftdetection. Furthermore, the cement production process at PT XYZ was found to be statistically capable, with a lower capability index (Ppl) and process performance index (Ppk) of 1.45, indicating consistent compliance with lower specification limits and centered process performance. These results affirm the practical effectiveness of the AEWMA ME C AI Max chart in enhancing process monitoring and capability assessment in industrial applications.

Downloads

Download data is not yet available.
Published
2025-04-30