Parameter Identification of Lithium-Ion Batteries Model to Predict Discharge Behaviors Using Heuristic Algorithm

Abstract

A complete set of parameters that describes lithium-ion batteries is required in order to better control the sophisticated battery management system used during practical application of lithium-ion batteries. Previous work in the literature requires long times and a computer cluster to identify the full set of parameters while using a pseudo-two-dimensional (P2D) model and a genetic algorithm (GA). This work proposes a more efficient parameter identification method based on the heuristic algorithm. The heuristic algorithm includes features such as the requirement of fewer identified parameters and the effective interval spacing of parameters to save searching time during the GA. This algorithm also uses a divide-and-conquer strategy that divides the full set of P2D parameters into two groups and identifies each group, respectively. With this method, the complete identification of P2D model parameters can be performed within 10 h using a single core computer. Simulated data using the above identified parameters are consistent with the reference results from the literature. And the identified paramters could be used for estimation of the state of charge and the state of available power capability of lithium-ion batteries.