Modelling fuel consumption and refuelling of autonomous vehicles

Zöldy, Máté; Zsombók, Imre · 2018 · OpenAlex-citations

DOI: 10.1051/matecconf/201823500037

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This paper addresses the under-researched problem of refuelling optimization for autonomous vehicles, specifically focusing on Plug-in Hybrid Electric Vehicles (PHEVs). While significant research exists on automotive cybersecurity and trajectory following, the authors argue that refuelling strategies remain a critical gap in the development of fully autonomous systems. The study aims to define the parameters necessary for simulating and testing an artificial intelligence-supported decision-making system that optimizes refuelling. By focusing on PHEVs, the research seeks to create a model that encompasses both internal combustion engine (ICE) and battery electric vehicle (BEV) dynamics, as PHEVs utilize both conventional fuel and external electricity. The methodology consists of a comprehensive literature review and categorization of factors influencing fuel consumption and refuelling logistics. The authors first analyze various drivetrains, identifying PHEVs as the optimal subject for modelling due to their dual-energy nature. They then systematically categorize fuel consumption influencers into four groups: vehicle-driven parameters (e.g., weight, maintenance, tire pressure), road-related parameters (e.g., surface type, rolling resistance), usage-related parameters (e.g., driver style, acceleration severity, gear selection), and ambient parameters (e.g., temperature, landscape profile). The paper further reviews existing methods for predicting fuel consumption, evaluating techniques such as Random Forest, Support Vector Machines, and Artificial Neural Networks, noting that sampling rates and data integration significantly affect prediction accuracy. Key findings highlight that fuel consumption is heavily influenced by a complex interplay of internal and external factors. The review identifies that aggressive driving, improper maintenance, and short trips significantly increase fuel usage, while road surface quality and ambient temperature also play measurable roles. In terms of prediction modelling, the authors note that while various machine learning techniques show comparable performance, integrating road segmentation and specific sampling times (e.g., 10-minute intervals) improves accuracy. Regarding refuelling, the study finds that for electric components, range uncertainty and charging time are major constraints compared to the negligible refuelling time of ICE vehicles. The authors conclude that effective refuelling models must account for pricing strategies, vehicle weight changes due to fuel load, and real-time environmental data to optimize both cost and efficiency. The significance of this work lies in its establishment of a foundational framework for developing intelligent refuelling systems for autonomous vehicles. By identifying the specific influencers and modelling requirements for PHEVs, the research provides a basis for creating realistic simulations that can be extended to pure ICE and BEV fleets. The authors emphasize that understanding these consumption dynamics is essential for optimizing route planning, reducing operational costs, and enhancing the reliability of autonomous driving systems. This study bridges the gap between theoretical fuel consumption models and practical refuelling logistics, offering a direction for future research in automated fleet management and eco-driving assistance.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-26; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• situational awareness