A machine learning framework for predicting fuel consumption and CO2 emissions in hybrid and combustion vehicles: comparative analysis and performance evaluation

Rania A Ibrahim and Nahla E Zakzouk

PLOS ONE, 2026, vol. 21, issue 2, 1-30

Abstract: Accurate estimations of fuel consumption and carbon emissions insights are critical for performance benchmarking, emissions compliance, and the optimization of energy management strategies in vehicles’ systems. Unlike model-based predictive approaches that require complex modelling, machine learning (ML) predictive models learn patterns directly from data, w making them flexible, automated, and scalable solutions for complex nonlinear systems that can easily adapt to diverse sets of data with high predictive accuracy. These models typically span from linear and nonlinear models to ensemble approaches, where the latter are often preferred owing to their ability to aggregate multiple learners and more effectively capture intricate relationships.. This study develops a predictive ML framework for estimating vehicle emissions and fuel consumption in lightweight vehicles via a real-world dataset. The primary contribution of this work lies in the fusion and integration of internal combustion engine vehicle (ICEV) and plug-in hybrid vehicle (PHEV) datasets into a common modelling workflow, whereas most existing studies rely solely on combustion-vehicle datasets only. Another contribution is the dual-forecast capability of the proposed model, enabling simultaneous prediction of both vehicle emissions and fuel usage rather than solely predicting emissions, as in most prior studies. In contrast, this study offers a unified framework capable of accurately forecasting both vehicle emissions and energy consumption. The adopted broader and more diverse mixed dataset enhances generalization, in addition to making the proposed model a practical and reliable tool for environmental assessment, sustainable vehicle development, and policy decision-making.

Date: 2026
References: Add references at CitEc
Citations:

Downloads: (external link)
https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0342418 (text/html)
https://journals.plos.org/plosone/article/file?id= ... 42418&type=printable (application/pdf)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:plo:pone00:0342418

DOI: 10.1371/journal.pone.0342418

Access Statistics for this article

More articles in PLOS ONE from Public Library of Science
Bibliographic data for series maintained by plosone ().