CHEMICAL ENGINEERING TRANSACTIONS VOL. 78, 2020 A publication of The Italian Association of Chemical Engineering Online at www.cetjournal.it Guest Editors: Jeng Shiun Lim, Nor Alafiza Yunus, Jiří Jaromír Klemeš Copyright © 2020, AIDIC Servizi S.r.l. ISBN 978-88-95608-76-1; ISSN 2283-9216 Moving Windows Prediction of Refined, Bleached and Deodorized Palm Oil Quality Using Multiple Least Squares and Partial Least Squares Regressions Wai Hoong Khu a , Nor Adhihah Rashid a , Mohd. Aiman Mohd. Noor a , Nur Atikah Mohd Rosely a , Azmer Shamsuddin c , Kamarul Asri Ibrahim b , Mohd. Kamaruddin Abd. Hamid b, * a Process Systems Engineering Centre (PROSPECT), Research Institute of Sustainable Environment (RISE), Universiti Teknologi Malaysia, 81310 UTM Johor Bahru, Johor, Malaysia b School of Chemical and Energy Engineering, Faculty of Engineering, Universiti Teknologi Malaysia, 81310 UTM Johor Bahru, Johor, Malaysia c Lahad Datu Edible Oils Sdn. Bhd., KM 2, Jalan Minyak off Jalan POIC, Locked Bag 16, 91109 Lahad Datu, Sabah, Malaysia kamaruddin@cheme.utm.my With the goal of shaping a smart, economically-sustainable palm oil refining industry, this paper aims to improve the prediction modelling of Refined, Bleached and Deodorized Palm Oil (RBDPO) quality for early quality fault detection and remediation via the novel use of a moving windows form of prediction in addition to the conventional static window form. In this study, both Multiple Least Squares Regression (MLSR) and Partial Least Squares Regression (PLSR) model training techniques and prediction computations are carried out in the form of two moving data windows types, namely the expanding and rolling fixed-size data windows, together with the conventional static window form of prediction as control. Prediction improvement is observed consistently across both regression techniques when carried out in moving windows form, where both types of moving windows predictions, i.e. expanding and rolling windows, have fared significantly better than the static window form, with an average prediction error reduction of 20.6 % for Free Fatty Acid prediction, 55.9 % for Moisture Content prediction, 32.6 % for Iodine Value prediction and 34.2 % for Colour Value prediction in RBDPO. Among the moving windows themselves, the superiority of expanding windows over rolling windows and vice versa is insignificant. On the whole, this study paves the way for a revamped RBDPO quality prediction form which better reflects the dynamic, transient nature of the palm oil refining process, thus further consolidating its reliability for widespread practical use. 1. Introduction While palm oil has been a commodity favored in oleochemical research, concerns have been raised regarding on the imbalance in research strategies since the efforts are geared towards technical topics in favor of sustainability issues in palm oil production (Hansen et al., 2015). This study is thusly indirectly oriented towards the economical and quality sustainability aspect of palm oil refining, in which the predictive modelling of RBDPO quality as explored here is expected to provide quality forecasts which allow operators to monitor product quality in advance and to perform early remediations of crude oil feed. If substandard RBDPO quality could be predicted and prevented, it will reduce the need for cost-and-energy intensive recycling processes, which according to Makky and Soni (2014), can amount up to an hourly opportunity loss of MYR 159,000 due to excessive energy dedicated to pumps and process equipment during re-processing. With improved economic performance, the sustainable development of palm oil refinery plants can be further encouraged, thus positively reinforcing the path towards a smarter, sustainable palm oil refining industry (Ngan et al., 2018). To a certain extent, the implementation of RBDPO quality prediction also caters to the environmental sustainability aspect, where scraps and waste can be minimized as the production of off-spec products can be DOI: 10.3303/CET2078074 Paper Received: 14/05/2019; Revised: 04/08/2019; Accepted: 28/08/2019 Please cite this article as: Khu W.H., Rashid N.A., Mohd. Noor M.A., Mohd Rosely N.A., Shamsuddin A., Ibrahim K.A., Abd. Hamid M.K., 2020, Moving Windows Prediction of Refined, Bleached and Deodorized Palm Oil Quality Using Multiple Least Squares and Partial Least Squares Regressions, Chemical Engineering Transactions, 78, 439-444 DOI:10.3303/CET2078074 439