Statistical Quality Control of NPK Fertilizer Production Process using Mixed Dual Multivariate Cumulative Sum (MDMCUSUM) Chart based on Multioutput Least Square Support Vector Regression (MLS-SVR)

DOI

10.2991/978-94-6463-332-0_2How to use a DOI?

Keywords

Control Charts; MDMCUSUM; MLS-SVR; NPK Fertilizer

Abstract

Each year, the price of non-subsidized fertilizer has increased. The price increases may reduce demand. However, quality also has an impact on demand. To maintain demand, the quality of non-subsidized fertilizer must be monitored. Using the Mixed Dual Multivariate Cumulative Sum (MDMCUSUM) chart, this study will monitor the quality of non-subsidized NPK (Nitrogen, Phosphorus, and Potassium) fertilizer. The MDMCUSUM chart integrates two different types of Multivariate Cumulative Sum (MCUSUM) charts into a single chart to detect a specific shift in the process mean. Two types of MDMCUSUM, the CP (CMCUSUM and PRMCUSUM) chart and the PC (PRMCUSUM and CMCUSUM) chart, will be used, and the performance will be compared with the assumption that the mean shift is [0.75, 1.5]. However, autocorrelation in the data has led to an increase in false alarms. To overcome the issue, the Multioutput Least Squares Support Vector Regression (MLS-SVR) model is used. The MLS-SVR algorithm, which utilizes RBF (Radial Basis Function) kernel functions and grid search methods to find the optimal hyperparameters, will be used to generate the residuals. Those residuals are then used to construct the MDMCUSUM chart. Using the optimal hyper-parameters, the MLS-SVR model has successfully reduced autocorrelation in the residuals. When the MDMCUSUM chart is used to monitor the residuals in phase I, out-of-control observations are detected. The optimal hyperparameters obtained from phase I are used in phase II. Further, in phase II, the process is already in-control. Based on monitoring results, the PC chart can detect out-of-control signals faster than the CP chart.

Copyright

© 2023 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Margaretha Gracia Hotmatua Silalahi
AU  - Muhammad Ahsan
AU  - Muhammad Hisyam Lee
PY  - 2023
DA  - 2023/12/18
TI  - Statistical Quality Control of NPK Fertilizer Production Process using Mixed Dual Multivariate Cumulative Sum (MDMCUSUM) Chart based on Multioutput Least Square Support Vector Regression (MLS-SVR)
BT  - Proceedings of the 5th International Conference on Statistics, Mathematics, Teaching, and Research 2023 (ICSMTR 2023)
PB  - Atlantis Press
SP  - 4
EP  - 13
SN  - 2352-538X
UR  - https://doi.org/10.2991/978-94-6463-332-0_2
DO  - 10.2991/978-94-6463-332-0_2
ID  - Silalahi2023
ER  -