Smart Energy Utilization for Electric Bus System Using Machine Learning

Abstract

The sudden increase in electric buses in city transportation has raised the need of efficient energy management and fleet optimization. Still the systems in use mainly concentrate on tracking and scheduling rather than smart ways of predicting energy use in various operating conditions. This paper proposes a machine learning approach for predicting energy use in electric bus systems and optimizing fleet operations. The proposed system uses several input factors such as vehicle speed, acceleration, battery charge level, temperature, traffic conditions, road type, and passenger load for accurate predictions of energy use in electric bus systems. A hybrid machine learning approach using ensemble methods has been proposed for high accuracy in predictions. Additionally, the system proposes route optimization by comparing various routes and choosing the most efficient route for energy use in electric bus systems. The system also provides insights into battery use, estimated travel distance. A web-based system has been proposed for real-time predictions and user interaction. The test results prove that this approach has high accuracy in predictions and significantly helps in reducing energy use and costs. The proposed system can support a sustainable and smart transportation system for electric bus systems.

Introduction

The study addresses the optimization of energy consumption in electric bus fleets using a machine learning-based predictive system. Electric buses offer environmental and operational advantages over diesel buses but face challenges due to energy usage variability caused by factors like traffic, driving style, weather, passenger load, and route characteristics. Existing public transport systems focus mainly on tracking and scheduling and lack real-time energy prediction or intelligent route optimization.

The proposed system analyzes operational parameters—including speed, acceleration, battery condition, traffic, weather, and passenger load—to predict energy consumption accurately using ensemble machine learning methods. It also performs route optimization, recommending energy-efficient paths and supporting battery and range management. With a web-based interface, the system facilitates real-time decision-making for fleet managers, improving operational efficiency, reducing costs, and enabling intelligent, data-driven urban transport planning.

Key contributions include integrating energy prediction and route optimization, accounting for dynamic real-world conditions, and providing a user-friendly interface for practical deployment.

Conclusion

This paper is a machine learning-based prediction system of energy use and optimal operation of the fleet of electric buses. The proposed method is an efficient measure to use several parameters of operation and environment to estimate the use of energy under the real-life circumstances. The system has been designed by combining an ensemble learning model with route optimization approaches to make exact predictions and find energy-saving routes. The invented solution can also provide other insights like battery consumption, approximate traveling range, and efficiency in operation which will help transport officials in making decisions. A web based interface further makes it easier to use since real time interaction and analysis are made possible. On the whole, the suggested system helps to decrease energy usage, decrease the operational cost, and enhance the efficiency of the electric bus systems. It helps in creating intelligent and sustainable public transportation that uses data-driven methods.

References

[1] T. Pamu?a and D. Pamu?a, “Prediction of Electric Buses Energy Consumption from Trip Parameters Using Deep Learning,” Energies, vol. 15, no. 5, pp. 1747, 2022. DOI: https://doi.org/10.3390/en15051747 [2] P. Wang et al., “Energy Consumption Estimation Method of Battery Electric Buses Based on Real-World Driving Data,” World Electric Vehicle Journal, vol. 15, no. 7, 2024. DOI: https://doi.org/10.3390/wevj15070314 [3] Y. Xing, Z. Chen, and X. Li, “Operation Energy Consumption Estimation Method of Electric Buses Based on CNN,” Mathematical Problems in Engineering, vol. 2022, Article ID 6904387, 2022. DOI: https://doi.org/10.1155/2022/6904387 [4] G. Zhu, “Energy Consumption Prediction Model for Electric Buses Based on Actual Operation Data,” Advances in Transportation Research, 2024. DOI: https://doi.org/10.1155/2024/3058575 [5] M. A. Rahman et al., “Predictive Modeling of Energy Demands for Battery Electric Bus Fleet Based on Large-Scale Real-World Data,” Energy Informatics, 2025.DOI: https://doi.org/10.1186/s42162-025-00564-y [6] U. Vijay et al., “Valuation of Public Bus Electrification with Open Data,” arXiv preprint arXiv:2209.12107, 2022. DOI: https://doi.org/10.48550/arXiv.2209.12107 [7] S. Zhang et al., “Extended Vehicle Energy Dataset (eVED): A Dataset for Electric Vehicle Energy Research,” arXiv preprint arXiv:2203.08630, 2022. DOI: https://doi.org/10.48550/arXiv.2203.08630 [8] G. S. Oh et al., “Vehicle Energy Dataset (VED): A Large-Scale Dataset for EV Energy Modeling,” arXiv preprint arXiv:1905.02081, 2019. DOI: https://doi.org/10.48550/arXiv.1905.02081 [9] Hendriks et al., “Integrated Electric Bus Energy Model and Depot Charging Optimization,” Nature Energy, 2024. DOI: https://doi.org/10.1038/s44333-024-00008-2 [10] Z. Wang et al., “Online Energy Consumption Forecast for Electric Buses Using Lightweight Models,” Scientific Reports, 2025. DOI: https://doi.org/10.1038/s41598-024-82432-5

Copyright

Copyright © 2026 Palreddy Srinivas Reddy, Ganta Priyanka, Midde Swathi, Mrs. C Maheswari. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.