Controlling Industrial Processes Using Multivariate Exponential Weighted Moving Average (Mewma)
Evi Ramadhani1, Herman Mawengkang2, Sutarman3, Marwan Ramli4

1Evi Ramadhani, Graduate Study, Department of Mathematics Program, Indonesia.
2Herman Mawengkang, Syiah Kuala University, Banda Aceh, Indonesia.
3Sutarman, North Sumatera University, Medan, Indonesia.
4Marwa Ramli, North Sumatera University, Medan, Indonesia.
Manuscript received on 09 May 2019 | Revised Manuscript received on 19 May 2019 | Manuscript Published on 23 May 2019 | PP: 1448-1453 | Volume-7 Issue-6S5 April 2019 | Retrieval Number: F12560476S519/2019©BEIESP
Open Access | Editorial and Publishing Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (

Abstract: This paper discusses the Multivariate Exponential Weighed Moving Average (MEWMA) in controlling industrial processes. MEWMA is a multivariate control diagram that can be used to detect autocorrelation and detect the mean vector shift. But the success of the measurement with the MEWMA control diagram depends on the smoothing parameter𝝀and the comparative weight𝝎. So far, it has never been specifically seen whether the MEWMA control chart can meet certain assumptions, but can provide appropriate information. Another multivariate control diagram that can be used to control the process is the T 2Hotelling control chart . Therefore, this initial phase will be simulated with the MEWMA control chart, and the results will be compared with the T 2Hotellingas a comparison. The results of the study showed that the MEWMA control chart yielded more sensitive results to detect the shift in the mean vector process compared to the multivariate T 2 -Hotelling control chart.
Keywords: T2 Hotelling, Parameter 𝝀, Comparison Weight𝝎, Mean Vector.
Scope of the Article: Industrial Engineering