Machine Learning Predictive Algorithm for Self-Sensing Electric Vehicle Battery Enclosure

Abstract Electric Vehicles (EVs) are a favorable and rapidly growing tactic for reducing carbon emissions. However, the most commonly used power source in EVs, Lithium-Ion Batteries (LIBs), can pose a significant safety risk in the form of thermal runaway. This is a fast-acting and dangerous failure mode that may lead to fires and explosions. To address this issue, the authors’ previous work developed a self-sensing composite battery enclosure with embedded micro-temperature sensors to provide LIB condition monitoring. The prior work produced extensive experimental and simulation results, characterizing an enclosure-embedded battery management system. It was found that the top composite layer causes a time lag in temperature detection, impeding an early warning signal. This current study aims to create a regression model leveraging machine learning (ML) strategies which is able to predict interior battery enclosure temperatures when trained on the prior study’s thermal experiments and simulations. The temperature inference model predicts the enclosure’s surface temperatures using embedded temperature measurements in real-time, compensating for the time lag. Random Forest (RF) and Recurrent Neural Network (RNN) ML models are compared, considering performance and computational costs. Mean Absolute Error (MAE) and Mean Absolute Percentage Error (MAPE) are utilized to quantify the prediction accuracies of both approaches. The temperature inference model enhances the practicality of a self-sensing composite battery enclosure as a battery management system, mitigating risks associated with LIB thermal runaway events. By monitoring embedded temperature changes and predicting the temperatures on the interior surface of the enclosure, the system provides insights into potential hazards, enabling timely interventions and ensuring EV safety.