Improving Fuzzy Statistical Clustering Approach for Estimating the Change in X-bar Control Chart
| Autor: | Wei-Jhih Lin, 林韋志 |
|---|---|
| Rok vydání: | 2015 |
| Druh dokumentu: | 學位論文 ; thesis |
| Popis: | 103 Statistical process control (SPC) charts are typically employed to monitor a process shift. If an out-of-control signal occurs in a SPC chart, engineers have to identity the cause and remove it. The method of identity the real time change is knows a change-point estimation problem. In practice, the shift type of a process is uncertain. Engineers need to identify the shift type. When the shift type is a step change, using a linear estimator results an inaccurate estimation. Similarly, if there exists a linear trend change type, using a step estimator yields a considerable delay. A slope coefficient of determination approach was proposed. When an out-of-control signal occurs, the current point and previous point are used to estimate the slope of the linear trend. According to the value of the estimated slope, the fuzzy statistical clustering approach was then used to identify the change point. From the simulation results and practice examples, the proposed approach enables effectively detecting step changes or linear trend changes in a process, thereby facilitating identifying the true process change point. |
| Databáze: | Networked Digital Library of Theses & Dissertations |
| Externí odkaz: |
|