Battery Thermal Model Identification and Surface Temperature Prediction

With the rapid growth of the electric and hybrid vehicles industry, the use of Lithium-ion (Li-ion) batteries has increased enormously in recent years. The modeling of Li-ion batteries helps in a better understanding of internal dynamics and prediction of their behavior. This results in the better development of a battery management system (BMS). The operation of Li-ion batteries in extreme hot and cold conditions negatively impacts their performance and life. Therefore, to overcome these problems and for better performance of batteries, thermal management is needed in batteries. This thesis aims at advancing the development of the thermal electrical equivalent circuit model (TECM) of batteries and the surface temperature prediction of batteries. State-of-the-art BTMS is designed to react to the measured temperature at the battery surface whereas research suggests that the performance of BTMS will likely improve when used in conjunction with the predicted temperature. A theoretical sound approach is presented in this thesis to model the heat generation in the core of the battery and its propagation to the surface. The least-square-based methods are developed for the estimation of model parameters such as battery internal resistance and thermal parameters. The performance of the developed parameter identication algorithm was tested at different signal-to-noise ratio (SNR) values. Experimental results using data collected from a high precision battery cycler and temperature sensor shows an open loop Normalized Mean Absolute Error (NMAE) of 6.17% or a maximum error of 3°C during a high-current discharge that lasted approximately 800 seconds. Since it is important to evaluate the performance of developed algorithms in real-world applications, a hardware-in-the-loop (HIL) system is developed for the verification of the designed temperature prediction algorithm.