RAHMAN, FATHUR (2017) MULTIVARIATE SHORT-RUN PRODUCTION CONTROL CHART FOR MONITORING VARIABILITY AND MEANS PROCESS. Masters thesis, Faculty of Mathematics and Science.
Preview |
Text
1311201011-Master_Theses.PDF - Published Version Download (4MB) | Preview |
Abstract
Statistical process control methods to monitor short-run production its considering by multivariate measurement, or the method is relatively new in control charts can also be called, multivariate control charts for short-run production to monitor the process mean and variability. To monitor the process mean control charts used influence function (IF) of mean and covariance matrix that is expected to detect small shifts, and to identify shifts in the variability of the process is used as the main component and eigenvalue influence function. The technique used is general, and the influence functions can be used to build multivariate control charts for both short-run nominal value or estimated. The methodology applied to the data creation subsea pipeline in the production PT.KHI (Krakatau Hoogeven International) in the 2014 period. A Results of the study showed the using conventional control charts can not detect a shift for variability and mean processes effectively, as is evident in the T2 control charts only two shifts that can be detected is the observation to 15 and 21 and on control charts for variables there are not any significant shifts can be detected, In the process of implementing the control charts for Short Production Run the shift of variability and mean the process can be detected with more leverage, which used influence functions (IF) of eigenvaluve and eigen vectors of the principal component for detecting a shift. in process variability and control chart T2 based influence functions (IF) of mean and matrix covariance of the principal component to detect the mean shift of the process by which each can detect three shifts simultaneously, in the observation to 15, 21 and 24.
=======================================================================================================
Metode statistik proses kontrol untuk memantau proses jangka pendek
(short-run) dengan mempertimbangkan pengukuran multivariat, atau metode
diagram kontrol yang terbilang baru ini bisa juga disebut, diagram kontrol
multivariat jangka pendek (short-run control chart) untuk memantau proses mean
dan variabilitas. Untuk memantau mean proses pada diagram kontrol digunakan
fungsi pengaruh mean dan matriks kovarian yang di harapkan dapat mendeteksi
pergeseran kecil, dan untuk mengetahui pergeseran pada variabilitas proses maka
digunakan komponen utama dan eigenvalue sebagai pengaruh fungsi. Teknik
yang digunakan bersifat umum, dan pengaruh fungsi dapat digunakan untuk
membangun diagram kontrol multivariat jangka pendek (short-run) baik untuk
nilai nominal atau estimasi. Metode ini lebih lanjut diterapkan pada data
pembuatan pipa bawah laut yang di produksi PT.KHI (Krakatau Hoogeven
International) pada periode 2014. Hasil dari penelitian menujukan penggunaan
diagram kontrol konvensional tidak dapat mendeteksi pergeseran variabilitas dan
mean proses secara efektif, ini terbukti pada diagram kontrol T2 hanya dua
pergeseran yang dapat dideteksi yaitu pada observasi ke 15 dan 21 sedangkan
pada diagram kontrol variabel tidak terdapat satu pun pergeseran yang signifikan.
Pada proses penerapan diagram kontrol Short Production Run pergeseran
variabilitas dan mean proses dapat di deteksi dengan lebih maksimal, dimana
digunakan fungsi pengaruh (IF) eigen valuve dan eigen vektor dari komponen
utama (principal component) untuk mendeteksi pergeseran pada variabilitas
proses dan diagram kontrol T2 berbasis fungsi pengaruh (IF) mean dan matriks
kovarian dari komponen utama untuk mendeteksi pergeseran mean proses dimana
dari masing-masing observasi dapat mendeteksi 3 pergeseran secara bersamaan
yaitu pada observasi ke 15, 21 dan 24.
| Item Type: | Thesis (Masters) |
|---|---|
| Uncontrolled Keywords: | Statistical Process Control, Short Run Production, Eigen value, Eigen vectors, Influence Function, statistik proses kontrol, Principal Component |
| Subjects: | H Social Sciences > HA Statistics Q Science Q Science > Q Science (General) Q Science > QA Mathematics |
| Divisions: | Faculty of Mathematics and Science > Statistics > 49101-(S2) Master Thesis |
| Depositing User: | Fathur Rahman |
| Date Deposited: | 25 Jan 2017 07:33 |
| Last Modified: | 06 Mar 2019 07:28 |
| URI: | http://repository.its.ac.id/id/eprint/3053 |
Actions (login required)
 |
View Item |