A Surrogate-Assisted Multi-Objective Framework for Thermal–Hydraulic Optimization of Serpentine Liquid Cooling in Lithium-Ion Battery Modules

Optimizing liquid-cooled battery modules using high-fidelity computational fluid dynamics (CFD) remains computationally demanding, particularly when multiple design variables and competing performance objectives must be considered simultaneously. To address this issue, this study develops a surrogate-assisted multi-objective framework for the thermal–hydraulic optimization of a serpentine liquid-cooling system for cylindrical lithium-ion battery modules. First, a three-dimensional CFD model is established and benchmarked against published results. Latin hypercube sampling is then used to generate representative design points, and Gaussian process regression (GPR) models are constructed to relate three design variables, namely channel wall thickness, coolant inlet velocity, and inlet temperature, to three performance metrics: maximum temperature, maximum temperature difference, and pressure drop. Morris global sensitivity analysis is further employed to quantify the relative importance of the design variables. The surrogate models are subsequently coupled with the Non-dominated Sorting Genetic Algorithm II (NSGA-II) to obtain Pareto-optimal solutions. The results indicate that inlet temperature has the strongest influence on the maximum temperature, inlet velocity predominantly affects temperature uniformity, and channel wall thickness mainly governs pressure drop. Compared with the baseline design, the selected Pareto solution reduces the pressure drop from 539.0 Pa to 208.8 Pa, corresponding to a reduction of 61.26%, while increasing the maximum temperature and maximum temperature difference by only 0.21% and 1.00%, respectively. A full CFD re-evaluation of the selected optimum yields relative errors below 2.5%, supporting the reliability of the proposed framework. The proposed approach provides a practical and computationally efficient tool for balancing thermal and hydraulic performance in liquid-cooled battery modules.