Nonparametric mixed exponentially weighted moving average-moving average control chart

Area of Science:

• Industrial Engineering
• Statistical Process Control
• Quality Management

Background:

• Traditional control charts often assume specific data distributions, limiting their applicability.
• Detecting subtle shifts in process location is crucial for maintaining product quality and efficiency.

Purpose of the Study:

• To design and evaluate a novel distribution-free control chart for identifying process location changes.
• To introduce a mixed exponentially weighted moving average-moving average (EWMA-MA) chart utilizing a signed-rank statistic.

Main Methods:

• Development of a distribution-free EWMA-MA control chart based on the signed-rank statistic.
• Utilizing Monte Carlo simulation techniques to generate the run-length profile of the proposed chart.
• Performance evaluation using symmetrical distributions and various individual/overall performance measures.

Main Results:

• The proposed EWMA-MA control chart demonstrates superior performance compared to existing charts in detecting process location shifts.
• The chart effectively utilizes recent and past sample information with differential weighting.
• Run-length profile analysis confirms the chart's enhanced sensitivity and efficiency.

Conclusions:

• The developed distribution-free EWMA-MA control chart offers a robust and effective tool for process monitoring.
• The chart provides a practical solution for quality control, as demonstrated by its application in a gas turbine setting.
• This research contributes a valuable, adaptable method to statistical process control literature.