Performance assessment of a novel hybrid cooling design for battery thermal management systems

Lithium-ion batteries' operating temperature range directly affects their longevity, safety, and performance. A considerable amount of heat is produced by the battery during operation. A heat management system is, therefore, essential to cool the battery. Air cooling is a widespread and acceptable technique to enhance the heat dispersion of battery thermal management (BTM) systems. Still, it is unsuccessful due to the limited thermal conductivity of air and the significant quantity of heat produced by the battery in a brief period of time. To investigate the viability of this strategy, four well-known air-cooling BTM systems—the Z-type, U-type, ZU-type, and Z-trapped-shaped BTM systems—are selected for study. Subsequently, four novel hybrid air-liquid cooling systems are proposed by studying the temperature nonuniformity of the cell, which has been overlooked in recent studies. Moreover, this study introduced the least square boosting (LSBoost) model, integrating Bayesian hyper-parameter optimization (BHPO) and 5-fold cross-validation for battery cell temperature prediction. According to the study's accomplishments, the hybrid ZU-shape-opt's maximum temperature (Tmax) and maximum temperature difference (ΔTmax) are 317.651 K and 18.583 K, respectively, which are 13.442 K and 3.81595 K (20.53 %) lower than the ZU-shape. The results showed that BHPO significantly ameliorated cell temperature prediction accuracy by identifying the optimal configuration of hyper-parameters for the proposed model.