Lithium-ion battery health state estimation based on improved snow ablation optimization algorithm-deep hybrid kernel extreme learning machine

State of health (SOH) estimation of lithium-ion batteries (LIBs) is crucial for battery health management systems. To accurately estimate the health state of LIBs, a hybrid model integrating improved snow ablation optimizer (ISAO), deep extreme learning machine (DELM), and hybrid kernel extreme learning machine (HKELM) is proposed. Firstly, health factors indicating battery degradation are extracted from the charge and discharge process curves of experimental data to provide an accurate description of the aging mechanisms of the battery. Secondly, the HKELM is incorporated into the regression layer of the DELM to yield the deep hybrid kernel extreme learning machine (DHKELM) model for estimating the SOH. Thirdly, the proposed ISAO algorithm mitigates the risk of the algorithm converging to a local optimum by integrating Latin hypercube sampling, Levy flight strategy, and normal cloud model. The critical parameters of the DHKELM model are optimized utilizing the ISAO algorithm. Finally, the proposed algorithm is verified by public data resources. The experimental results show that the proposed method exhibits higher accuracy and robustness for SOH estimation compared to other models.