Author ORCID Identifier
0000-0002-7913-820X
Standing
Graduate (Masters)
Type of Proposal
Poster Presentation
Faculty Sponsor
Dr. Balakumar Balasingam
Proposal
Over the past 30 years, a lot of advancement and progress have been received by Lithium-ion batteries. Due to this advancement, the use of Li-ion batteries increased in electric vehicles (EVs). Li-ion batteries are highly sensitive to operating conditions and temperature is one of them. The performance of Li-ion batteries is highly affected by temperature such as low temperature decreases the power output and high temperature degrades the health of batteries. To overcome these issues, proper thermal management of batteries is an important task. This work presents a thermal-electrical equivalent circuit model (TECM) for the prediction of surface temperature of batteries. In order to accurately study the thermal behavior of batteries using TECM model, the precise estimation of model parameters is an important task. This study shows the three methods for the estimation of model thermal parameters i.e. unconstrained least square (ULS), constrained least square (CLS), and weighted least square (WLS). The estimation performances of these algorithms are first shown using computer simulations at different values of signal-to-noise ratio (SNR) in order to consider the effect of different noise levels. WLS methods outperform the order two approaches in the estimation of thermal parameters. Later, these estimated parameters are used for the future prediction of surface temperature of real Li-ion batteries. This predictive approach is helpful in reliable thermal management of batteries. This advanced knowledge can be used to alter the thermal management system's settings, such as the increase/decrease of coolant flow rate, resulting in accurate temperature control.
Battery Thermal Model Identification and Surface Temperature Prediction
Over the past 30 years, a lot of advancement and progress have been received by Lithium-ion batteries. Due to this advancement, the use of Li-ion batteries increased in electric vehicles (EVs). Li-ion batteries are highly sensitive to operating conditions and temperature is one of them. The performance of Li-ion batteries is highly affected by temperature such as low temperature decreases the power output and high temperature degrades the health of batteries. To overcome these issues, proper thermal management of batteries is an important task. This work presents a thermal-electrical equivalent circuit model (TECM) for the prediction of surface temperature of batteries. In order to accurately study the thermal behavior of batteries using TECM model, the precise estimation of model parameters is an important task. This study shows the three methods for the estimation of model thermal parameters i.e. unconstrained least square (ULS), constrained least square (CLS), and weighted least square (WLS). The estimation performances of these algorithms are first shown using computer simulations at different values of signal-to-noise ratio (SNR) in order to consider the effect of different noise levels. WLS methods outperform the order two approaches in the estimation of thermal parameters. Later, these estimated parameters are used for the future prediction of surface temperature of real Li-ion batteries. This predictive approach is helpful in reliable thermal management of batteries. This advanced knowledge can be used to alter the thermal management system's settings, such as the increase/decrease of coolant flow rate, resulting in accurate temperature control.