A WAVELET-BASED VARIABLE CONTROL PROCEDURE FOR DETECTING PROCESS MEAN SHIFT

Authors

  • Chien-Chih Wang Professor
  • Wei-Min Wen

DOI:

https://doi.org/10.23055/ijietap.2010.17.3.82

Keywords:

Qulaity Control Chart

Abstract

This paper develops a wavelet-based approach for a variable control chart, and adopts the data decomposition and linear combination techniques to detect process shifts. The Shewhart, exponentially weighted moving average (EWMA), and cumulative sum (CUSUM) control charts are the most popular monitoring process graph tools. However, these charts were developed for different process situations. If a user chooses an inappropriate control chart to monitor a process, the correct control result will not be obtained. This study used the wavelet transform to develop a novel variable control procedure. First, the Haar function was used as the basis for data decomposition. Next, the linear combination technique was used to combine different resolution data through wavelet transform decomposition. Simulations were adopted to evaluate performance. An analysis showed that the detection ability of the wavelet-based variable control chart was superior to the EWMA control chart in a comparison of average run length (ARL) results.

Author Biographies

Chien-Chih Wang, Professor

Department of Industrial Engineering and Management

Wei-Min Wen

Department of Industrial Engineering and Management

Downloads

Published

2010-09-09

How to Cite

Wang, C.-C., & Wen, W.-M. (2010). A WAVELET-BASED VARIABLE CONTROL PROCEDURE FOR DETECTING PROCESS MEAN SHIFT. International Journal of Industrial Engineering: Theory, Applications and Practice, 17(3). https://doi.org/10.23055/ijietap.2010.17.3.82

Issue

Section

Quality, Reliability, Maintenance Engineering