A Multilevel Hyperheuristic Optimization Algorithm Framework for Online Parametrization of an Electrochemical Impedance Model on Lithium-Ion Batteries

The electrochemical impedance spectrum (EIS) is recognized as an effective battery experimental measure, which has been confirmed for online engineering practice according to the chip-level on-board battery management system. However, for the application of battery state estimation, the parametrization for an electrochemical impedance model is necessary along with the state degradation of batteries. However, the conventional fitting methods suffer from the imbalance for searching precision and calculational cost. In this article, a method constructed by a multilevel swarm-intelligent hyperheuristic optimization algorithm is proposed, which integrates the algorithms including the fireworks algorithm, particle swarm optimization, osprey optimization algorithm, and Harris hawks optimization. The proposed algorithm takes the advantages of different heuristic optimization algorithms with specific performances on searching field and speed and can be further extended according to the actual requirements. Eight common optimization functions and various types of batteries are used for validating the proposed multilevel algorithm, including a 50 Ah NCM power battery, a 97 Ah LFP power battery, and a parallel-connected 600 Ah energy storage battery. The results show that the proposed method performs with satisfactory convergence under different states of charge and degradation conditions. The article contributes to solving the issue of parameter identification of online EIS samples and can be further extended to other fields such as equivalent circuit model identification with the flexible adjustment to other swarm-intelligent optimization methods according to the requirements.