Heat dissipation optimization of lithium-ion battery pack based on neural networks. (5th November 2019)

• Record Type: Journal Article
• Authors: Qian, Xiao; Xuan, Dongji; Zhao, Xiaobo; Shi, Zhuangfei
• Is Part Of: Applied thermal engineering. Volume 162(2019)
• Publication Date: 2019-11-05
• Subjects: Computational fluid dynamics -- Forced air cooling -- Bayesian neural network -- Battery spacing -- Heat dissipation optimization; Heat engineering -- Periodicals; Heating -- Equipment and supplies -- Periodicals; Periodicals; 621.40205
• DOI: 10.1016/j.applthermaleng.2019.114289
• Languages: English
• ISSNs: 1359-4311

Abstract

Highlights: Studied the battery temperature within the battery pack using ANSYS Fluent. Studied the influence of battery spacings on cooling performance of battery pack. The optimization variable number was reduced by grouping the batteries. The Bayesian neural network model was established with CFD simulation results. Cooling effect of battery pack was improved by adjusting the battery spacings. Abstract: The excessively high temperature of lithium-ion battery greatly affects battery working performance. To improve the heat dissipation of battery pack, many researches have been done on the velocity of cooling air, channel shape, etc. This paper improves cooling performance of air-cooled battery pack by optimizing the battery spacing. The computational fluid dynamics method is applied to simulate the flow field and temperature field of the battery pack for different battery spacing. The battery spacing and corresponding CFD simulation outputs (maximum temperature and temperature difference) are used to train the Bayesian neural network. Compared with CFD simulation results, the relative errors are 0.08% and 3.2%. With this neural network model, the optimal battery spacing arrangement is found which is [17, 24, 22, 0.22, 0.23, 0.176, 0.176] and the temperature difference and the maximum temperature of the batteries are respectively 5.986(K) and 300.511(K). The results show this neural network model can accurately describe the relationship between the battery spacing and the battery temperature. This optimization process represents an effective and time-saving method to design the battery spacing distribution to improve the cooling performance of battery pack.