Diagram Kontrol Adaptive Exponentially Weighted Moving Average Max (AEWMAM) Chart

Andikaputra, Salman Alfarizi Pradana (2023) Diagram Kontrol Adaptive Exponentially Weighted Moving Average Max (AEWMAM) Chart. Masters thesis, Institut Teknologi Sepuluh Nopember.

Text 6003211011-Master_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 October 2025. Download (5MB) | Request a copy

Abstract

Statistical Process Control (SPC) adalah teknik statistik yang banyak digunakan untuk memastikan bahwa suatu proses memenuhi kriteria tertentu. SPC digunakan untuk mengawasi standar, melakukan pengukuran, dan mengambil tindakan korektif dalam produksi. Diagram kontrol variabel merupakan diagram kontrol yang digunakan untuk mengukur karakteristik kualitas yang terukur dengan mengendalikan mean proses dan varians proses. Diagram kontrol univariat adalah diagram kontrol yang menggunakan satu karakteristik kualitas. Diagram kontrol shewart dengan data yang diamati pada waktu saat ini mampu mendeteksi adanya pergeseran proses mean atau variasi yang besar, sedangkan untuk mendeteksi adanya pergeseran proses mean atau variasi yang kecil yakni menggunakan diagram kontrol Cumulative Sum (CUSUM) dan diagram kontrol Exponentially Weighted Moving Average (EWMA). Disisi lain, diagram kontrol EWMA juga mengalami pengembangan. Mengingat diagram kontrol EWMA hanya bekerja dengan baik pada pergeseran mean yang kecil namun tidak pada pergeseran besar. EWMAM lebih unggul daripada diagram simultan lainnya yang ada yakni dengan mendeteksi pergeseran kecil dalam parameter proses, hal ini juga baik dalam mendeteksi pergeseran besar. Metode Adaptive EWMA yang merupakan kombinasi dari diagram EWMA dan skema Shewhart dengan memodifikasi smoothing parameter yang dimiliki sehingga mampu memberikan performa terbaik pada pergeseran kecil maupun besar. Skema AEWMA lebih unggul dari EWMA konvensional dalam pengurangan banyak efek buruk dalam keadaan tetap, dan evaluasi beberapa grafik menggunakan ukuran tetap. Proses monitoring dilakuan secara parsial yakni pada mean dan variabilitas. Pertama variabilitas terlebih dahulu dimonitor selanjutnya memonitor mean. Hal tersebut memakan waktu yang lebih banyak dan kurang efektif karena harus membuat terlebih dahulu dua diagram dengan satu persatu. Berdasarkan hasil analisis diagram kontrol AEWMAM bekerja sedikit lebih baik dibanding diagram kontrol EWMAM dan Max-chart pada pergeseran kecil baik secara mean, varians, dan mean varians, selain itu pada pergeseran besar baik secara mean, varians, dan mean varians pada diagram kontrol AEWMAM lebih baik daripada EWMAM namun masih belum lebih baik dari diagram kontrol Max-chart. Kinerja Diagram Kontrol AEWMAM memberikan hasil yang cukup optimal untuk mendeteksi adanya sinyal out-of-control pada parameter λ=0.1 dengan subgrup n=5. Hal ini ditandai dengan nilai ARL1 yang dihasilkan semakin cepat mendekati nilai 1 pada λ=0.1 dan subgrup n=5 berdasarkan pergeseran mean,varians, dan mean varians daripada lambda dan subgrup yang lain. Selain itu pada penerapan data langsung hasil monitoring pada data suhu pada tambak udang vannamei CV Air Mas Jaya dengan menggunakan diagram kontrol AEWMAM mampu memberikan hasil yang cukup optimal dengan mendeteksi sebanyak mungkin sinyal out-of-control dan meminimalisir terjadinya false alarm
=================================================================================================================================
Statistical Process Control (SPC) is a statistical technique widely used to ensure that a process meets certain criteria. SPC is used to oversee standards, take measurements, and take corrective action in production. Variable control chart is a control chart used to measure measurable quality characteristics by controlling the mean process and process variance. A univariate control chart is a control chart that uses a single quality characteristic. Shewart control chart with data observed at the current time is able to detect a shift in the mean process or a large variation, while to detect a shift in the mean process or a small variation of the Cumulative Sum (CUSUM) control chart and control chart Exponentially Weighted Moving Average (EWMA). On the other hand, EWMA control charts are also undergoing development. Considering the EWMA control chart only works well on small mean shifts but not on large shifts. EWMAM is superior to other existing simultaneous diagrams by detecting small shifts in process parameters, it is also good at detecting large shifts. Adaptive EWMA method which is a combination of EWMA diagram and Shewhart scheme by modifying smoothing parameters that are owned so as to provide the best performance on small and large shifts. The AEWMAM scheme is superior to conventional EWMA in the reduction of many adverse effects in a fixed state, and the evaluation of multiple graphs using a fixed size. The monitoring process is carried out partially on the mean and variability. First the variability is first monitored then monitor the mean. This is more time consuming and less effective because you have to create two diagrams one by one. Based on the analysis of AEWMAM control chart works a little better than EWMAM control chart and Max-chart on small shifts both in the mean, variance, and mean variance, in addition to large shifts both in the mean, variance, and mean variance on AEWMAM control chart is better than EWMAM but still not better than Max-chart control chart. The performance of AEWMAM control chart provides optimal results to detect the presence of out-of-control signal in parameter λ =0.1 with subgroup n=5. This is characterized by the value of ARL1 that is produced faster and closer to the value of 1 at the α=0.1 and n=5 subgroups based on the shift of the mean,variance, and mean variance than lambda and other subgroups. In addition to the application of direct data monitoring results on temperature data on shrimp farms vannamei CV Air Mas Jaya using AEWMAM control chart is able to provide optimal results by detecting as many out-of-control signals and minimizing the occurrence of false alarms

Item Type:	Thesis (Masters)
Uncontrolled Keywords:	Adaptive Exponentially Weighted Moving Average Max-Chart (AEWMAM), Diagram Kontrol, Statistical Process Control (SPC), Univariat, Suhu, Vannamei, Control Chart, Temperature, Univariate
Subjects:	H Social Sciences > HA Statistics H Social Sciences > HA Statistics > HA29 Theory and method of social science statistics H Social Sciences > HA Statistics > HA31.38 Data envelopment analysis. H Social Sciences > HD Industries. Land use. Labor > HD9980.5 Service industries--Quality control.
Divisions:	Faculty of Science and Data Analytics (SCIENTICS) > Statistics > 49101-(S2) Master Thesis
Depositing User:	Salman Alfarizi Pradana Andikaputra
Date Deposited:	13 Sep 2023 07:43
Last Modified:	13 Sep 2023 07:43
URI:	http://repository.its.ac.id/id/eprint/104531

Actions (login required)

View Item