Recent research led by Mahmoud M. Kiasari from the Smart Grid and Green Power Systems Research Laboratory at Dalhousie University has unveiled promising advancements in fire protection for electric vehicle (EV) batteries. As the automotive industry shifts towards electric vehicles, the safety of lithium-ion (Li-ion) batteries has become a pressing concern due to the risk of thermal runaway—an uncontrolled temperature rise that can lead to fires or explosions. This study, published in the journal ‘Fire’, explores the integration of Thermal Energy Storage (TES) systems with machine learning (ML) techniques to improve battery safety and reliability.
The research specifically addresses the high-voltage battery systems used in EVs during Vehicle-to-Grid (V2G) and Grid-to-Vehicle (G2V) operations. These modes involve complex interactions between the vehicle and the electrical grid, which can exacerbate thermal management challenges. By employing TES, which can absorb and release thermal energy, the study aims to mitigate the risks associated with overheating in these batteries.
Kiasari explains, “The combination of TES and ML offers a novel way to overcome challenges in thermal management. While TES provides physical solutions, ML delivers predictive capabilities to address potential thermal issues.” This dual approach allows for enhanced monitoring and management of battery conditions under various environmental scenarios.
Through simulations conducted in Simulink, the research analyzed how TES can effectively control battery temperatures. The findings revealed that the State of Charge (SoC) of the battery is the most critical factor influencing thermal management, while grid power consumption has a lesser impact. The study also highlighted that logistic regression outperformed other machine learning models in predicting thermal risks, achieving high accuracy and precision.
The implications for the energy sector are significant. As EV adoption grows, ensuring the safety of battery systems is paramount for consumer confidence and regulatory compliance. The integration of TES and ML could lead to safer battery designs, which may enhance the overall marketability of electric vehicles. Additionally, this technology could open up new business opportunities in the development of advanced energy storage solutions and smart grid technologies.
Looking ahead, Kiasari suggests that future research should focus on real-world implementations and collaborations with industry partners, including EV manufacturers. By expanding the scope of studies to include various energy consumption scenarios and advanced machine learning techniques, the potential for improved battery safety and efficiency could be realized.
This innovative research not only addresses critical safety concerns but also paves the way for a more sustainable and reliable future for electric vehicles, ultimately benefiting both consumers and the energy sector at large.
