Published Paper Details:

The growing adoption of Electric Vehicles (EVs) has led to an increasing need for advanced methods to monitor and predict battery performance in real-time. Accurate prediction of battery states, such as State of Charge (SoC) and State of Health (SoH), is critical for enhancing battery longevity, improving energy efficiency, and ensuring the reliability of EVs. This project presents a novel approach that integrates Machine Learning and Digital Twins to predict battery states in EVs using data-driven machine learning techniques. The proposed model utilizes real-time data collected from EV battery systems to create a digital replica, or "digital twin," which simulates the behaviour of the battery over time. The system is trained using machine learning algorithms, including regression and classification models, to predict key battery parameters. An essential feature of this approach is its emphasis on explainability, ensuring that the machine learning model's decisions are interpretable and transparent to users and engineers. This is achieved through the use of interpretable models and feature importance analysis, allowing stakeholders to understand how different factors affect battery performance. The model is implemented in Python using popular machine learning libraries, with real-time data input from EV Battery Management Systems (BMS). Experimental results demonstrate the effectiveness of the approach in accurately predicting battery states while maintaining model transparency. This work provides a promising framework for the development of intelligent, data-driven systems that can optimize the performance and lifespan of EV batteries, ultimately contributing to the sustainability and efficiency of electric mobility.