A Joint SOH-SOC Estimation Method Based on BiLSTM for Lithium-ion Batteries

• Published in: 2025 2nd International Conference on Electrical Technology and Automation Engineering (ETAE) pp. 264 - 268
• Main Authors: Qi, Jiajin, Yu, Shujiang, Chen, Yunlong, Xu, Yinfei, Zhu, Jiangyuxuan, Li, Jiangtao
• Format: Conference Proceeding
• Language: English
• Published: IEEE 23.05.2025
• Subjects: Accuracy; Bidirectional long short term memory; BiLSTM; Estimation; Feature extraction; Joint estimation; Lithium batteries; Lithium-ion batteries; NASA; Real-time systems; Robustness; State of charge; State of health; The incremental capacity approach
• DOI: 10.1109/ETAE65337.2025.11089775

Accurate and efficient the state of health (SOH) and state of charge (SOC) of estimation is critical for improving the reliability of energy storage systems and promoting their application in new power grids. Current research primarily aims to improve SOH and SOC accuracy while neglecting the mutual influence of SOH and SOC across different time scales. This paper presents a BiLSTM-based method for joint prediction of SOH and SOC. Initially, the constant current charging process of lithium batteries is examined, and key features are extracted using the incremental capacity approach. Then, the interrelation between SOH and SOC is explored, and an estimation model is constructed using the BiLSTM network. Finally, SOH estimation is performed for a battery under a certain operating condition from the National Aeronautics and Space Administration (NASA) lithium battery dataset, and SOC estimation is carried out for the battery state under a specific SOH.