Optimal thermal management of electric vehicle battery systems using serpentine minichannel cold plates with intersecting V-shaped minichannels

Environmental concerns are promoting the shift towards electric vehicles (EVs) from internal combustion-based vehicles. Lithium ion (Li-ion) batteries are currently the dominant power source option for modern electric vehicles (EVs); however, their operating temperatures need to remain within an allowable safety range in order to preserve battery lifetime and avoid thermal runaway. Accordingly, high-performance battery thermal management systems (BTMSs) are needed for safe and efficient battery operation. These challenges are addressed here using a novel machine learning (ML)-enabled multi-objective optimization (MOO) approach for BTMS based on serpentine minichannel cold plates with intersecting V-shaped minichannels (SMCCP-IVSMC). The SMCCP-IVSMC configuration is optimised here for the first time, subject to four competing objectives, namely the battery maximum temperature ( T max ), water pumping power ( P w ), battery temperature standard deviation ( T σ ), and the mass of the cold plate ( M CP ). Reducing M CP has not been considered previously; however, it plays a vital role in improving both the energy consumption and manufacturing costs of cooling systems. A thermal model based on empirical heat generation with a conjugate heat transfer model in the cold plates is developed and validated comprehensively. Surrogate modeling based on both Radial Basis Functions (RBFs) and Gaussian Process Regression (GPR) shows the latter is the most effective. This is used to explore the trade-offs between the competing objectives ( T max , P w , T σ , and M CP ). A novel hybrid optimization approach is developed, integrating GPR, Generalized Differential Evaluation (GDE3) and Technique for Order Preference by Similarity to an Ideal Solution (TOPSIS) methods to determine the best compromise solutions (BCSs) among the set of Pareto optimal solutions. The optimisation results identify numerous options for improving performance significantly, beyond the current benchmark design. In addition to reductions in the maximum temperature, operating costs can be minimised by reducing water pumping power by over 68.7 %, and manufacturing costs minimised by reducing the mass of the cold plate heat exchangers by over 22.8 %. The top-ranked BCS using equally weighted TOPSIS has also been identified, which enables all four objectives to be reduced simultaneously.