A Hybrid Data-Driven Method for State-of-Charge Estimation of Lithium-Ion Batteries
Xiaodong Yan,Gongbo Zhou,Wei Wang,Ping Zhou,Zhenzhi He
DOI: https://doi.org/10.1109/jsen.2022.3188845
IF: 4.3
2022-08-15
IEEE Sensors Journal
Abstract:As a portable energy storage system, lithium-ion batteries (LIBs) are widely used in wireless sensor networks, electric vehicles and other fields. To ensure the continuity of power supply, it is necessary to monitor the state of charge (SOC) of LIBs. However, due to the nonlinearity of battery operation, accurate SOC estimation has become a challenging task. In this paper, a SOC estimation method based on long-term short-term memory (LSTM) network and improved particle filter (IPF) is proposed, which maps the easily observed voltage, current and temperature to the target SOC. Firstly, through a layer of the LSTM network, the timing characteristics of the data are fully utilized to obtain the SOC variation trend of LIBs. Then, the noise variance adaptive algorithm and particle distribution optimization algorithm are introduced to improve the standard particle filter (PF). On this basis, the estimation results of the LSTM network are optimized by IPF. In addition, the performance of the proposed LSTM-IPF method is compared with other methods. The results show that the estimation performance of the proposed model is excellent, and the root mean squared error (RMSE) and maximum error (MAX) are controlled below 1% and 2% respectively, which meets the requirements of SOC estimation and verifies the effectiveness of the proposed method.
engineering, electrical & electronic,instruments & instrumentation,physics, applied