A Comparative Study on Parameter Identifications of Battery Thermoelectric Coupling Model

In this paper, two equivalent thermal models exhibiting the thermal effects of lithium-ion (Li-ion) batteries are considered------first order and second-order equivalent thermal models. In order to obtain high accuracy of the parameters of the electronic components representing the temperature in the thermal models, the recursive least squares (FFRLS) with forgetting factor and the proposed recursive least squares algorithm based on adaptive forgetting factor and discount factor (VWF-VFFRLS) and the coyote algorithm (COA) are used to identify the parameters of the thermal model. The results show that for the first-order thermal model, the difference between the three is not significant. However, for the second-order thermal model, the wolf algorithm based on off-line identification has higher accuracy than the other two on-line identification methods. The parameters of the battery thermoelectric coupling model are identified under the condition of complete charge and discharge of the battery. Through verification, it is found that the surface temperature error RMSE after parameter verification obtained by the wolf algorithm is 0.1217, which is much lower than 0.2512 of FFRLS and 0.2441 of VFFRLS.