Incorporating driving behavior into vehicle fuel consumption prediction: methodology development and testing
DOI: 10.1007/s43621-024-00511-z
archive: archived pipeline: cataloged verified
Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)
Summary
This study addresses the challenge of accurately predicting vehicle fuel consumption by integrating driving behavior into predictive models. While driving style significantly impacts fuel efficiency, existing methods often lack integrated frameworks that combine behavioral metrics with environmental factors for real-time prediction. The authors propose a comprehensive, multi-step methodology to link driving volatility measures to fuel consumption, aiming to provide actionable insights for improving fuel economy and supporting the development of Advanced Driver Assistance Systems (ADAS). The research utilizes naturalistic driving data from the inD dataset, collected in urban mixed-traffic environments in Germany. The dataset comprises trajectories for approximately 5,408 road users, including cars, trucks, buses, pedestrians, and cyclists, recorded over 10 hours at four intersections. The methodology involves several key steps: first, extracting ten specific volatility measures (such as standard deviations and coefficients of variation for speed and acceleration) to characterize driving behavior. Second, applying an unsupervised K-means clustering algorithm to classify drivers into three categories: conservative, normal, and aggressive. Third, estimating instantaneous fuel consumption using the Virginia Tech Comprehensive Power-based Fuel Consumption Model (VT-CPFM), a microscopic physics-based model that calculates fuel use based on vehicle power, mass, and resistance forces. Finally, the study develops predictive models using these behavioral features and fuel consumption data, comparing a baseline Multivariate Linear Regression model against advanced machine learning techniques. The results demonstrate that machine learning approaches significantly outperform traditional statistical methods. The baseline linear regression model achieved an R-squared value of 0.511 and a Mean Squared Error (MSE) of 0.031, indicating moderate predictive accuracy. In contrast, the Random Forest model achieved a substantially higher R-squared of 0.956 and a lower MSE of 0.003, confirming its superior ability to capture the non-linear relationships between driving behavior and fuel consumption. Furthermore, the analysis revealed that conservative and aggressive driving behaviors lead to significantly higher and more variable fuel consumption compared to normal driving. Smooth, controlled driving was identified as the most fuel-efficient style. The significance of this work lies in its provision of a robust framework for real-time fuel consumption prediction that accounts for driver behavior. By demonstrating the high accuracy of Random Forest models in this context, the study supports the integration of such algorithms into fleet management systems and ADAS. The findings suggest that promoting normal driving habits can yield substantial fuel savings and environmental benefits. This methodology advances the field by bridging the gap between microscopic driving dynamics and macroscopic fuel efficiency predictions, offering a practical tool for optimizing vehicle routing and encouraging sustainable transportation practices.
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.
| Stage | Outcome | Tool | Model | Prompt | Attempts | Completed |
|---|---|---|---|---|---|---|
| discover | success | OpenAlex-citations | — | — | 1 | 2026-06-20 |
| archive | success | unpaywall | — | — | 2 | 2026-06-26 |
| extract | success | cached | — | — | 2 | 2026-06-26 |
| clean | success | clean | — | — | 1 | 2026-06-20 |
| chunk | success | chunk | — | — | 1 | 2026-06-20 |
| embed | success | embed | Qwen/Qwen3-Embedding-8B | — | 1 | 2026-06-20 |
| promote | success | — | — | — | 1 | 2026-06-20 |
| summarize | success | llm | qwen3.6-27b-prismaquant | summ-v5 | 1 | 2026-06-26 |
| tag | success | vector_similarity | — | — | 6 | 2026-06-20 |
| verify | success | — | — | — | 1 | 2026-06-26 |
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.
Information type
What kind of knowledge this paper contributes, grouped by family — independent of topic (what it is about) and method (how it was studied).
- Theoretical Contribution: computational model