Benchmarking Linear vs. Non-Linear Predictive Models for Energy Demand Forecasting in Electrical Motor-Driven Compressors

Authors

Humam Nazir
Technical University of Malaysia Malacca image/svg+xml
Tarmizi Ahmad Izzuddin
Technical University of Malaysia Malacca image/svg+xml
Nur Izyan Zulkafli
Technical University of Malaysia Malacca image/svg+xml
Mohamad Fani Bin Sulaima
Technical University of Malaysia Malacca image/svg+xml
Mohd Hafiz Jali
Technical University of Malaysia Malacca image/svg+xml
Haslenda Hashim
Nur Saleha Jayiddin
Azmi Md Lasin
M Tarmidzi Iskandar
DOI: https://doi.org/10.52549/ijeei.v14i1.7777
Download PDF
Published: 18-03-2026 • Issue: Vol. 14 No. 1: March 2026 • Section: Regular paper

Abstract

In the LNG regasification plants, among the most energy-demanding components are the electrical motor driven compressors. Artificial Neural Networks (ANNs) and Extreme Gradient Boosting (XGBoost) feature superior flexibility in modelling nonlinear and dynamic compressor behaviour, unlike the conventional methods such as Multiple Linear Regression (MLR), which is limited by its assumption of linearity. This study focuses on data-driven techniques for compressor power consumption prediction, in which MLR, ANN and XGBOOST are compared. Here, real operational data collected from two Boil-off Gas (BOG) and two Regasification Terminal Export Compressors (RGTEC) were considered to develop and evaluate the models. The findings suggest that non-linear machine learning models provide superior predictive performance in comparison to those of the traditional linear approach. The highest prediction accuracy for most compressors is achieved using the ANN model, with a n R2 values of 94.6%, 99.64% and 99.64% for BOG-A, BOG-C and RGTEC-A, respectively. Meanwhile, the best performance was found for XGBOOST for RGTEC-B, with a R2 value of 94.17% and a significantly lower RMSE value. Contrary to that, MLR produced lower accuracy in several cases, particularly when subjected to complex operating conditions, in which linear assumptions were inadequate to capture system dynamics. In addition, this research features one of the first direct comparisons between linear and nonlinear models applied to real-time compressor data, unlike past research that depends on simulations or single-method analysis. It emphasizes the practical advantages of ANN and XGBoost for data-driven energy forecasting and operational optimizations in the gas industry.
Keywords: Load Forecasting; Deep Learning; Multiple Linear - Regression; Electrical Compressors

References

M. S. Khan, M. A. Qyyum, W. Ali, A. Wazwaz, K. B. Ansari, and M. Lee, “Energy Saving through Efficient BOG Prediction and Impact of Static Boil-off-Rate in Full Containment-Type LNG Storage Tank,†Energies 2020, Vol. 13, Page 5578, vol. 13, no. 21, p. 5578, Oct. 2020, doi: 10.3390/EN13215578.

G. Kang, J. Im, and C.-J. Lee, “Operational Strategy to Minimize Operating Cost in Lng Terminal Using a Comprehensive Boil-Off Gas Model,†2023, doi: 10.2139/SSRN.4642375.

Z. Bouabidi et al., “Study on boil-off gas (Bog) minimization and recovery strategies from actual baseload lng export terminal: Towards sustainable lng chains,†Energies (Basel)., vol. 14, no. 12, Jun. 2021, doi: 10.3390/EN14123478.

U. Humme, P. Radgen, S. Ülker, and R. Schelle, “Findings from Measurements of the Electric Power Demand of Air Compressors,†Energies 2021, Vol. 14, Page 8395, vol. 14, no. 24, p. 8395, Dec. 2021, doi: 10.3390/EN14248395.

P. Sommani, S. Yoomak, and S. Ananwattanaporn, “Estimation of Power Consumption in Screw Gas Compressor Installed on Natural Gas Power Plant,†pp. 101–110, 2024, doi: 10.1007/978-3-031-67987-2_10.

D. Gryboś and J. S. Leszczyński, “A Review of Energy Overconsumption Reduction Methods in the Utilization Stage in Compressed Air Systems,†Energies 2024, Vol. 17, Page 1495, vol. 17, no. 6, p. 1495, Mar. 2024, doi: 10.3390/EN17061495.

M. Mukherjee et al., “Greenhouse gas emissions from the US liquefied natural gas operations and shipping through process model based life cycle assessment,†Communications Earth & Environment 2025 6:1, vol. 6, no. 1, pp. 1–14, Jan. 2025, doi: 10.1038/s43247-024-01988-2.

C. D. Davis et al., “Community Health Impacts From Natural Gas Pipeline Compressor Stations,†Geohealth, vol. 7, no. 11, p. e2023GH000874, Nov. 2023, doi: 10.1029/2023GH000874.

A. K. Chaudhari, “Integrating machine learning with industrial automation for enhanced predictive maintenance,†Using Computational Intelligence for Sustainable Manufacturing of Advanced Materials, pp. 487–509, Apr. 2025, doi: 10.4018/979-8-3693-7974-5.CH021.

L. Zhao et al., “Multi-Objective Optimization of Gas Storage Compressor Units Based on NSGA-II,†Energies 2025, Vol. 18, Page 3377, vol. 18, no. 13, p. 3377, Jun. 2025, doi: 10.3390/EN18133377.

M. Gambelli et al., “The Influence of Variable Operating Conditions and Components on the Performance of Centrifugal Compressors in Natural Gas Storage Reservoirs,†Energies 2025, Vol. 18, Page 3930, vol. 18, no. 15, p. 3930, Jul. 2025, doi: 10.3390/EN18153930.

W. Zhang, Q. Tang, X. Gao, and R. Zhou, “Multi-Period Operational Modelling and Optimization for Large-Scale Natural Gas Networks Considering Linepack Functions in Long-Distance Transmission Pipelines,†Journal of Marine Science and Engineering 2025, Vol. 13, Page 201, vol. 13, no. 2, p. 201, Jan. 2025, doi: 10.3390/JMSE13020201.

Y. Liu, Y. Li, G. Li, Y. Lin, R. Wang, and Y. Fan, “Review of multiple load forecasting method for integrated energy system,†Front. Energy Res., vol. 11, p. 1296800, Nov. 2023, doi: 10.3389/FENRG.2023.1296800/BIBTEX.

L. Zhong, R. Liu, X. Miao, Y. Chen, S. Li, and H. Ji, “Compressor Performance Prediction Based on the Interpolation Method and Support Vector Machine,†Aerospace 2023, Vol. 10, Page 558, vol. 10, no. 6, p. 558, Jun. 2023, doi: 10.3390/AEROSPACE10060558.

Q. Peng et al., “Centrifugal compressor performance prediction and dynamic simulation of natural gas hydrogen blended,†Int. J. Hydrogen Energy, vol. 52, pp. 872–893, Jan. 2024, doi: 10.1016/J.IJHYDENE.2023.10.023.

“Ai In The Energy Industry Statistics: Reports 2025.†Accessed: Sep. 08, 2025. [Online]. Available: https://wifitalents.com/ai-in-the-energy-industry-statistics/

D. Belayadi, D. Bendouda, N. Benali, and N. Hamouche, “Selecting an Appropriate Ensemble Machine Learning Algorithm for Intelligent Quality Management in Production,†Communications in Computer and Information Science, vol. 2606 CCIS, pp. 211–225, 2025, doi: 10.1007/978-3-032-01922-6_18.

Z. Rais, S. Muhammad Fahmuddin, Saida, and A. T. Utomo, “Implementation of Machine Learning Algorithm with Extreme Gradient Boosting (XGBoost) Method in Hypertension Level Classification,†Journal of Applied Science, Engineering, Technology, and Education, vol. 7, no. 1, pp. 126–136, Apr. 2025, doi: 10.35877/454RI.ASCI4191.

P. Jyotheeswari, N. Basyal, S. P. Gaire, and C. Patel, “Evaluating the impact of gamma (γ) and lambda (λ) on regression using XGBoost regressor,†Digital Transformation and Sustainability of Business, pp. 404–407, Jan. 2025, doi: 10.1201/9781003606185-93.

“What is regasification? | Clarksons.†Accessed: Aug. 27, 2025. [Online]. Available: https://www.clarksons.com/glossary/regasification/

“High Pressure BOG Compressor | Mitsui E&S Group.†Accessed: Aug. 28, 2025. [Online]. Available: https://www.mes.co.jp/en/products/bog/

“pandas - Python Data Analysis Library.†Accessed: Aug. 28, 2025. [Online]. Available: https://pandas.pydata.org/

D. Singh and B. Singh, “Investigating the impact of data normalization on classification performance,†Appl. Soft Comput., vol. 97, p. 105524, Dec. 2020, doi: 10.1016/J.ASOC.2019.105524.

K. Mahmud Sujon, R. Binti Hassan, Z. Tusnia Towshi, M. A. Othman, M. Abdus Samad, and K. Choi, “When to Use Standardization and Normalization: Empirical Evidence from Machine Learning Models and XAI,†IEEE Access, vol. 12, pp. 135300–135314, 2024, doi: 10.1109/ACCESS.2024.3462434.

“StandardScaler — scikit-learn 1.7.1 documentation.†Accessed: Aug. 28, 2025. [Online]. Available: https://scikit-learn.org/stable/modules/generated/sklearn.preprocessing.StandardScaler.html

“MinMaxScaler — scikit-learn 1.7.1 documentation.†Accessed: Aug. 28, 2025. [Online]. Available: https://scikitlearn.org/stable/modules/generated/sklearn.preprocessing.MinMaxScaler.html

G. Ciulla and A. D’Amico, “Building energy performance forecasting: A multiple linear regression approach,†Appl. Energy, vol. 253, p. 113500, Nov. 2019, doi: 10.1016/J.APENERGY.2019.113500.

M. Lai, D. Zhang, Y. Li, X. Wu, and X. Li, “Application of Multiple Linear Regression and Artificial Neural Networks in Analyses and Predictions of the Thermoelectric Performance of Solid Oxide Fuel Cell Systems,†Energies 2024, Vol. 17, Page 4084, vol. 17, no. 16, p. 4084, Aug. 2024, doi: 10.3390/EN17164084.

“1.1. Linear Models — scikit-learn 1.7.1 documentation.†Accessed: Aug. 29, 2025. [Online]. Available: https://scikit-learn.org/stable/modules/linear_model.html#ordinary-least-squares

O. A. Montesinos López, A. Montesinos López, and J. Crossa, “Fundamentals of Artificial Neural Networks and Deep Learning,†Multivariate Statistical Machine Learning Methods for Genomic Prediction, pp. 379–425, 2022, doi: 10.1007/978-3-030-89010-0_10.

M. G. M. Abdolrasol et al., “Artificial Neural Networks Based Optimization Techniques: A Review,†Electronics 2021, Vol. 10, Page 2689, vol. 10, no. 21, p. 2689, Nov. 2021, doi: 10.3390/ELECTRONICS10212689.

T. Szandała, “Review and comparison of commonly used activation functions for deep neural networks,†Studies in Computational Intelligence, vol. 903, pp. 203–224, 2021, doi: 10.1007/978-981-15-5495-7_11/TABLES/6.

D. P. Kingma and J. L. Ba, “Adam: A Method for Stochastic Optimization,†3rd International Conference on Learning Representations, ICLR 2015 - Conference Track Proceedings, Dec. 2014, Accessed: Aug. 31, 2025. [Online]. Available: https://arxiv.org/abs/1412.6980v9

N. D. Hoang, “Image Processing-Based Spall Object Detection Using Gabor Filter, Texture Analysis, and Adaptive Moment Estimation (Adam) Optimized Logistic Regression Models,†Advances in Civil Engineering, vol. 2020, no. 1, p. 8829715, Jan. 2020, doi: 10.1155/2020/8829715.

“3.1. Cross-validation: evaluating estimator performance — scikit-learn 1.7.1 documentation.†Accessed: Aug. 31, 2025. [Online]. Available: https://scikit-learn.org/stable/modules/cross_validation.html#cross-validation

“train_test_split — scikit-learn 1.7.1 documentation.†Accessed: Aug. 31, 2025. [Online]. Available: https://scikitlearn.org/stable/modules/generated/sklearn.model_selection.train_test_split.html

D. Chicco, M. J. Warrens, and G. Jurman, “The coefficient of determination R-squared is more informative than SMAPE, MAE, MAPE, MSE and RMSE in regression analysis evaluation,†PeerJ Comput. Sci., vol. 7, pp. 1–24, Jul. 2021, doi: 10.7717/PEERJ-CS.623/SUPP-1.

J. Hunt, “Pandas and Data Analytics,†pp. 611–627, 2023, doi: 10.1007/978-3-031-40336-1_54.

M. Rigacci, R. Sato, and K. Shirase, “Power Consumption Simulation of Servo Motors Focusing on the Influence of Mechanical Vibration on Motor Efficiencyâ€, doi: 10.20965/ijat.2022.p0104.

D. Hu, J. Liu, F. Yi, Q. Yang, and J. Zhou, “Enhancing heat dissipation to improve efficiency of two-stage electric air compressor for fuel cell vehicle,†Energy Convers. Manag., vol. 251, p. 115007, Jan. 2022, doi: 10.1016/J.ENCONMAN.2021.115007.

Y. Gebreyesus, D. Dalton, D. De Chiara, M. Chinnici, and A. Chinnici, “AI for Automating Data Center Operations: Model Explainability in the Data Centre Context Using Shapley Additive Explanations (SHAP),†Electronics (Switzerland), vol. 13, no. 9, May 2024, doi: 10.3390/ELECTRONICS13091628.

How to Cite

[1]
H. Nazir et al., “Benchmarking Linear vs. Non-Linear Predictive Models for Energy Demand Forecasting in Electrical Motor-Driven Compressors”, IJEEI, vol. 14, no. 1, pp. 135–152, Mar. 2026, doi: 10.52549/ijeei.v14i1.7777.