Biofuel consumption and emission prediction for harbour craft using Machine learning methods

Qingyao LI, Jasmine Siu Lee LAM,
Biofuel consumption and emission prediction for harbour craft using Machine learning methods,
Transportation Research Part D: Transport and Environment,
Volume 149,
2025,
105005,
ISSN 1361-9209,
https://doi.org/10.1016/j.trd.2025.105005.
(https://www.sciencedirect.com/science/article/pii/S1361920925004158)
Abstract: Biofuel represents a promising alternative fuel for reducing port environmental impacts. This study applies and compares various machine learning (ML) techniques, including Artificial Neural Networks (ANN), Long Short-Term Memory (LSTM), Ridge Regression (RR), Bayesian Ridge (BR), Extreme Gradient Boost (XGBoost), Random Forest (RF), and Support Vector Machines (SVM), in predicting biofuel consumption, emissions of carbon monoxide (CO), carbon dioxide (CO2), nitrogen oxides (NOx), and blending ratio classification for harbour craft in Singapore. Data was collected by conducting real sea trials. Results demonstrate that RF delivers superior performance in predicting fuel consumption and emissions, while SVM excels in biofuel blending ratio classification. The predicted results further reveal that biofuel blends slightly increase fuel consumption and onboard NOx emissions but reduce CO2 and CO emissions. These findings provide insights for stakeholders to optimise fuel management, emission control, and regulatory compliance in port operations.
Keywords: Biofuel; Fuel consumption prediction; Emission prediction; Ship emission; Machine learning; Harbour craft