Kamboj Rahul Kumar, Singh Mukesh, Singh Ashima, Bala Anju
Electrical and Instrumentation Engineering, Thapar Institute of Engineering and Technology, Street, Patiala, Punjab, 147004, India.
Computer Science and Engineering, Thapar Institute of Engineering and Technology, Street, Patiala, Punjab, 147004, India.
Sci Rep. 2025 Sep 1;15(1):32190. doi: 10.1038/s41598-025-17145-4.
In recent years, electric vehicles (EVs) have become increasingly popular, driven by advancements in battery technology, growing environmental awareness, and the demand for sustainable transportation. Compared to internal combustion engines, EVs not only produce fewer emissions but also offer greater energy efficiency, leading to reduced operating costs. Despite these advantages, concerns about battery failures have been a significant safety issue for EVs. This paper introduces an Intelligent Fault Detection (IFD) system-a proactive approach that utilises advanced intelligent techniques for detecting faults in EVs batteries. This paper involves developing and implementing an ML-based fault detection mechanism to monitor and safeguard the batteries from various faults, including thermal protection, under-voltage, and over-voltage. This research involves real-time data sourced from sensors embedded within the battery, which ensures the continuous collection of battery data during the vehicle's operation. The pre-processed data is analysed using a K-means clustering algorithm to classify essential groups, helping to define a valid range for temperature and voltage. Further, the ensemble approach has been used to classify safe and unsafe areas for battery operations. The proposed model has 0.94 accuracy for identifying faults, which contributes to the long-term sustainability and economic viability of electric mobility.
近年来，在电池技术进步、环保意识增强以及对可持续交通需求的推动下，电动汽车（EV）越来越受欢迎。与内燃机相比，电动汽车不仅排放更少，而且能源效率更高，从而降低了运营成本。尽管有这些优点，但电池故障问题一直是电动汽车的一个重大安全问题。本文介绍了一种智能故障检测（IFD）系统——一种利用先进智能技术检测电动汽车电池故障的主动方法。本文涉及开发和实施一种基于机器学习的故障检测机制，以监测和保护电池免受各种故障影响，包括热保护、欠压和过压。这项研究涉及从嵌入电池内的传感器获取实时数据，这确保了在车辆运行期间持续收集电池数据。使用K均值聚类算法对预处理后的数据进行分析，以对基本组进行分类，有助于定义温度和电压的有效范围。此外，集成方法已被用于对电池运行的安全和不安全区域进行分类。所提出的模型识别故障的准确率为0.94，这有助于电动出行的长期可持续性和经济可行性。
