A Hybrid Machine Learning and Physics-based Approach for Accurate Energy Consumption Modeling of Electric Buses in Public Transport

Aug 2, 2026

Upload by Rowan

Public transport organizations are increasingly focused on reducing air pollution, driving fleet transitions toward electric vehicles. Limited battery range and charging times create a need for accurate energy consumption modeling, especially under real-world variability from driving patterns and external influences. This study develops a hybrid model that combines physical principles with machine learning and leverages real telemetry data from 30 buses across 130 routes over one year.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages16

Reading Time40 min

AccessFree to download

Introduction

Motivation and challenges

Related work and modeling approaches

Rule-based vs data-driven methods

Methodology

Hybrid model design

Data and key variables

Machine learning models compared

Results and evaluation

Error metrics (MAPE)

Test vs validation performance

Impact of operational factors

| | | | | --- | --- | --- | | Document en libre accès dans PolyPublie
Open Access document in PolyPublie
| | | --- | --- | | URL de PolyPublie:
PolyPublie URL: | [https://publications.polymtl.ca/65988/](https://publications.polymtl.ca/65988/) | | Version officielle de l'éditeur / Published version
Version:
Révisé par les pairs / Refereed
| | | Conditions d’utilisation:
Terms of Use: | Creative Commons Attribution 4.0 International (CC BY) | | Document publié chez l’éditeur officiel
Document issued by the official publisher
| | | | | --- | --- | --- | --- | | Titre de la revue:
Journal Title: | | Sensors & Transducers (vol. 268, no. 1) | | |
Maison d’édition:
International Frequency Sensor Association
Publisher:

| | | | | | URL officiel:
Official URL:
| [https://www.proquest.com/scholarly-journals/hybrid-machine-learning](https://www.proquest.com/scholarly-journals/hybrid-machine-learning)physics-based-approach/docview/3 212840245/se-2?accountid=40695 | | | Mention légale:
Legal notice: | | © 2025. This work is published under [https://creativecommons.org/licenses/by/4.0/](https://creativecommons.org/licenses/by/4.0/)(the"License") . Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.
| | Ce fichier a été téléchargé à partir de PolyPublie, le dépôt institutionnel de Polytechnique Montréal This file has been downloaded from PolyPublie, the institutional repository of Polytechnique Montréal [https://publications.polymtl.ca](https://publications.polymtl.ca) Sensors & Transducers, Vol. 268, Issue 1, April 2025, pp. 45-58 Sensors & Transducers Published by IFSA Publishing, S. L., 2025 [http://www.sensorsportal.com](http://www.sensorsportal.com) A Hybrid Machine Learning and Physics-based Approach for Accurate Energy Consumption Modeling of Electric Buses in Public Transport 1,2,3 * Lucas ADAM, 1,3,4 Robert PELLERIN, and 1,2,3 Bruno AGARD 1 Polytechnique Montréal-Department of Mathematical and Industrial Engineering, Canada 2 LID, Data Intelligence Laboratory, Polytechnique Montréal, Canada 3 CIRRELT-Interuniversity Research Centre on Enterprise Networks, Logistics and Transportation, Canada 4 Jarislowsky/AtkinsRéalis Research Chair in the Management of International Projects, Canada * E-mail: lucas.adam@polymtl.ca Received: 12 Nov. 2025 /Revised:31March 2025 /Accepted:15 April 2025 /Published:30 April 2025 Abstract: Public transport organizations are increasingly concerned about reducing air pollution, leading many to transition their fleets into electric vehicles (EVs) . In this context, limited battery range and charging times remain significant hurdles. Precise modeling of electric bus energy consumption is crucial. Still, existing methods often face difficulties due to the complexities of real-world conditions, such as diverse driving patterns and external factors. To tackle this, the study proposesa hybrid model combining physical principles and machine learning using real-world data from 30 buses across 130 routes over one year. Key variables like passenger load, weather, and route characteristics are incorporated. Several machine learning models, including MLP, KAN, and XGBoost, are compared using Mean Absolute Percentage Error (MAPE) . The hybrid model outperforms others, achieving a low MAPE of 5.59 % on test data and 5.79 % on validation data with a low Standard Deviation. Additionally, models incorporating operational factors, such as bus lines and time of day, enhance prediction accuracy. The study concludes that integrating physical laws with machine learning offers a more accurate and stable approach to energy consumption modeling, providing a promising framework for fleet management and energy efficiency in public transport systems. Keywords: Public transport, Electric bus, Telemetry, Energy consumption, Big data, Artificial inte

FAQs

Why is accurate electric bus energy consumption modeling important?

Electric buses face constraints from limited battery range and longer charging times. Precise modeling supports more reliable fleet management and energy efficiency planning.

How does the proposed study improve on existing approaches?

It combines physical principles with machine learning in a hybrid model, using real-world telemetry data to better handle complex driving cycles and external factors.

Which variables and machine learning methods are used for prediction?

Key inputs include passenger load, weather, and route characteristics. The study compares models such as MLP, KAN, and XGBoost using Mean Absolute Percentage Error (MAPE).

Aug 2, 2026

Upload by Rowan