Development Of Two New Auxiliary Information Control Charts, And Economic And Economic-Statistical Designs Of Several Auxiliary Information Control Charts

• Main Author: Ng Peh, Sang
• Format: Thesis
• Language: English
• Published: 2020
• Subjects: QA1 Mathematics (General)
• Online Access: http://eprints.usm.my/54829/1/NG%20PEH%20SANG%20-%20TESIS%20cut.pdf; http://eprints.usm.my/54829/

The use of auxiliary information (AI) concept in control charts is receiving increasing attention among researchers. Control charts with auxiliary characteristics have been shown to be more efficient than control charts without such characteristics. The salient feature of the AI concept has motivated us to develop two new AI charts. The first objective of this thesis is to develop the run sum X - AI (RS X - AI) chart for monitoring the process mean. Optimal parameters computed using the optimization algorithms developed and the step-by-step approach for constructing the optimal RS - AI chart are provided in this thesis. The average run length (ARL) and expected average run length (EARL) performance criteria are used to evaluate the performance of the RS X - AI chart. Results show that the RS X - AI chart generally surpasses the existing X - AI, synthetic X - AI and EWMA X - AI charts in the detection of outof- control signals. The second objective of this thesis is to develop the variable sampling interval exponentially weighted moving average t AI (VSI EWMA t - AI) chart for monitoring the process mean when errors in estimating the process standard deviation exist. The VSI EWMA t - AI chart allows either the short or long sampling interval to be adopted, based on information from the process quality given by the current plotting statistic of the chart.