GRAFIK PENGENDALI MULTIVARIATE MODIFIED EWMA UNTUK OBSERVASI BERAUTOKORELASI (Studi Kasus: Pengendalian Kualitas Minyak Bumi pada Proses Feed Platforming PT. Pertamina Persero Refinery Unit II Dumai) MULTIVARIATE MODIFIED EWMA CONTROL CHART FOR AUTOCORRELATED OBSERVATIONS; (Case Study: Oil Quality Control of Feed Platforming Process in PT Pertamina Persero Refinery Unit II Dumai)

Mawadah, Ninda a, Herni Utami

2015 | Skripsi | FMIPA

The quality of product is one of factor which has big influence for consumer stratification. For maintain the quality, the products must be controlled in order to have product with good quality. The characteristics of quality consist of more than one variable which can be controlled by multivariate analysis. To control the characteristics usually used control chart. Be sides to control characteristics of industry quality, control chart can be used to control process. In many industries or processes there are some characteristics which have auto correlated observations like chemical process, paper and etc. for example is feed plat forming process. Feed plat forming process is one of process in oil industry which has purpose to change oil become big octane. To control the process can used Multivariate Modified EWMA control chart. Multivariate Modified EWMA is modification of multivariate EWMA which has function to control small shift and to identify abrupt change. In this mini thesis, use ARL (Average run length) to identify optimal Multivariate modified EWMA control chart with has definite mean shift. By comparing MMOEWMA control chart and MEWMA control chart, MMOEWMA control chart can detect abrupt change for autocorrelated observations.

Kata Kunci : MEWMA; Multivariate Modified EWMA; average run length; abrupt change; autocorrelation