Online Program Home
My Program

Abstract Details

Activity Number: 192
Type: Contributed
Date/Time: Monday, August 1, 2016 : 10:30 AM to 12:20 PM
Sponsor: Quality and Productivity Section
Abstract #319134
Title: A Double EWMA Control Chart for Individual Measurements Based on a Linear Prediction
Author(s): Rafael Perez Abreu* and Jay R. Schaffer
Companies: and University of Northern Colorado
Keywords: EWMA ; DEWMA ; Control Chart ; Trends ; Linear Prediction
Abstract:

Industrial process quality control frequently uses the exponential weighted moving average control chart (EWMA CC) and the double EWMA CC (DEWMA CC) to detect small shifts in a process. The EWMA CC was initially developed and evaluated in 1992. In 2005, the EWMA technique was extended to the DEWMA. Continued research into DEWMA has developed and assessed several alternatives, including multivariate control charts. These studies are focused on detecting small shifts in process. In practice, however, we occasionally wish to detect small trends, instead of shifts, in the process. The effectiveness of these methods to determine small trends a process is currently unknown. This research proposes a new control chart based on the fundamental theorem of exponential smoothing prediction, first presented by Brown (1960). This new chart is named "The Double Exponential Weighted Moving Average based on a Linear Prediction" (DEWMALP) control chart. We present a simulation study to contrast the efficiency between DEWMALP, EWMA, DEWMA, and classical Shewhart control charts when small trends are introduced. The conclusions about these contrasts are presented and implications are discussed.


Authors who are presenting talks have a * after their name.

Back to the full JSM 2016 program

 
 
Copyright © American Statistical Association