EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

A Physics-Based Normalized Reduced-Order Model for Trip-Level Energy Consumption Estimation of Electric Buses

Sakawrat Sricharoen, Komsan Hongesombut

IEEE Access · 2026 · 14 · 121916-121931

DOI: 10.1109/access.2026.3721214 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Access, 2026. Verfasst von Sakawrat Sricharoen und 1 weiteren Autor:innen. Die Arbeit nennt 40 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Reliable energy consumption (EC) estimation is critical for electric bus (EB) planning because it directly affects battery sizing, fleet allocation, charging infrastructure requirements, and operational feasibility. Existing physics-based models are interpretable but require high-resolution driving data and detailed vehicle parameters, whereas data-driven models depend on representative training datasets and often exhibit limited interpretability and transferability. This paper proposes a physics-based normalized reduced-order model for trip-level EB EC estimation using only three readily available trip-level inputs: average speed, total vehicle mass, and ambient temperature. Unlike reduced-order approaches that directly fit EC values, the proposed model references each estimate to a baseline operating condition and represents the effects of these variables through dimensionless correction factors associated with the dominant physical drivers of traction and HVAC energy demand. This baseline-referenced structure preserves physical interpretability, clarifies the leading-order contribution of each variable, and enables the transfer to new conditions by adjusting only the variables that change. The model was evaluated through simulation under different vehicle configurations, masses, temperatures, and driving profiles, and was validated using 70 real-world trips from an urban bus route in Bangkok, achieving a MAPE of 4.71% and an $R^{2}$ of 0.931. Compared with the five data-driven models, the proposed model achieved competitive accuracy with fewer input requirements and without target-route training. A planning case study demonstrated how trip-level EC estimates support battery sizing, fleet sizing, charger requirement assessment, and charging schedule feasibility analysis. Overall, the proposed model is suitable for preliminary EB planning on relatively flat urban routes, whereas high-fidelity simulation remains necessary for detailed design and operation-specific analyzes.

Zitieren

APA

Sricharoen, S., & Hongesombut, K. (2026). A Physics-Based Normalized Reduced-Order Model for Trip-Level Energy Consumption Estimation of Electric Buses. IEEE Access, 14, 121916-121931. https://doi.org/10.1109/access.2026.3721214

Vancouver

Sricharoen S, Hongesombut K. A Physics-Based Normalized Reduced-Order Model for Trip-Level Energy Consumption Estimation of Electric Buses. IEEE Access. 2026;14:121916-121931. doi:10.1109/access.2026.3721214

BibTeX
@article{Sricharoen2026,
  author = {Sricharoen, Sakawrat and Hongesombut, Komsan},
  title = {A Physics-Based Normalized Reduced-Order Model for Trip-Level Energy Consumption Estimation of Electric Buses},
  journal = {IEEE Access},
  year = {2026},
  volume = {14},
  pages = {121916--121931},
  doi = {10.1109/access.2026.3721214}
}
Kostenlos starten →