Lithium-Ion Battery Health State Estimation Based on Hierarchical Feature Construction and Adaptive Gated Sequence Network

State of health (SOH) estimation is vital in ensuring lithium-ion battery management systems' safe and stable operation (BMS). Data-driven methods are widely used to estimate SOH; however, existing methods often suffer from fixed or excessively high feature dimensions, impacting the model's adjustability and applicability. In this regard, this paper first proposes a layered knee-point strategy based on the charging voltage curve, which reduces the complexity of feature extraction. Then, a new hybrid framework called the Adaptive Gated Sequence Network (AGSN) model was proposed. The model integrates the Independent Recurrent Neural Network (IndRNN) layers, Active State Tracking Long Short-Term Memory (AST-LSTM) layers, and Adaptive Gating Mechanism (AGM) layers. By integrating a multi-layer structure and an adaptive gating mechanism, the SOH prediction performance was significantly improved. Finally, batteries under different operating conditions were tested using the NASA battery dataset. The results show that the AGSN model demonstrates higher accuracy and robustness in battery SOH estimation, with estimation errors consistently within 1%.