Estimation of thermal effect of different busbars materials on prismatic Li-ion batteries based on artificial neural networks

YETİK Ö., Karakoc T. H.

Journal of Energy Storage, vol.38, 2021 (SCI-Expanded) identifier identifier

  • Publication Type: Article / Article
  • Volume: 38
  • Publication Date: 2021
  • Doi Number: 10.1016/j.est.2021.102543
  • Journal Name: Journal of Energy Storage
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Compendex, INSPEC
  • Keywords: Numerical thermal analysis, Forced air cooling, Artificial neural network, Lithium-ion battery module, PHASE-CHANGE MATERIALS, MANAGEMENT-SYSTEM, MODULE, PERFORMANCE, COMPOSITE, PACK
  • Eskisehir Osmangazi University Affiliated: Yes


© 2021 Elsevier LtdThe need for energy is incrementally growing with the increasing population and developing industry. The reserves of fossil fuels such as diesel fuel and gasoline are close to depletion, and the environmental factors of these fuels are also negative. For these reasons, the shift from alternative energy sources to electric vehicles is accelerating. The most important parameter in the design of electric vehicles is the storage capacity of this energy. The heat generated during the charging and discharging of batteries used in these vehicles affects the lifespan and performance of the batteries. In addition, in an increasingly competitive environment, it is vital to obtain this information correctly in a short time. In this study, the way battery modules are affected by changing the busbar material of 10 series-connected prismatic batteries, different air velocity, and different air temperature values were evaluated and estimation was made by using Artificial Neural Networks (ANN) to reach the correct data in a short time. Silver, nickel, and steel were chosen as busbar materials. The best result was achieved with a silver (due to keeping the battery module at the lowest temperature). Evaluations were made at 1.0-1.5-2.0 and 2.5 C ratios, 1-2 m/s air inlet speeds, and 295-300 K air inlet temperatures. The estimation of the data was also carried out with the help of ANN. ANN algorithms BR-LM-CGP and SCG algorithms were evaluated. The best algorithm BR-16 was found. The R2 value of BR-16 was 0.995886, the CoV value was 0.005168 and the RMSE value was 0.011295.