Structural design and optimization of air-cooled thermal management system for lithium-ion batteries based on discrete and continuous variables

The power battery thermal management system plays a crucial role in controlling battery pack temperature and ensuring efficient battery operation. The optimal design of the structure of the battery thermal management system can greatly improve its thermal performance. The purpose of this paper is to address situations where structural parameters may exist as discrete or continuous variables, and to provide a more comprehensive design approach for similar battery thermal management systems. In this paper, different design optimization methods are adopted for different structural design variables. By comparing the implementation difficulty, stability and manufacturing cost, and thermal performance of the optimized battery pack model, the most suitable battery cooling system is determined. First, impact degree tests are conducted on the air inlet channel angle, side inclination angle, and battery cell spacing. These three parameters are considered discrete as random variables, and the significant factors and optimal parameters are determined using orthogonal technology and range analysis. Then, these three parameters are treated as continuous random variables, and the variables are sampled using the Latin hypercube sampling method. The functional relationship between the design variables and the target is constructed using the response surface method, and the multi -islands genetic algorithm is used for optimization. By comparing the structural models designed using both discrete and continuous parameter approaches, the difference between the maximum temperature and maximum temperature difference of the two heat dissipation systems is only 0.09 degrees C. After considering the continuous treatment of parameters, the calculation process is complicated, involving a large amount of data and interference terms, and the stability of the constructed model is worse than that of the discrete variables. Furthermore, the structure size designed using continuous variables has higher manufacturing accuracy and higher processing and manufacturing costs. Therefore, the thermal model designed using discrete variables is chosen. The results show that the maximum temperature and maximum temperature difference of the cooling system using the discrete parameter method are reduced by 5.5 % and 27.7 %, respectively.