A Novel Parameter Identification Method for SOC Estimation of Lithium-Ion Batteries based on Spatiotemporal Coupling Strategy

Accurately identifying equivalent circuit model (ECM) parameters is crucial for modeling and state estimation of lithium-ion batteries (LIBs). However, this task presents significant challenges due to the complex coupling effects of temperature and state of charge (SOC). To address this issue, we propose a novel parameter identification method for LIB based on spatiotemporal coupling strategy to enhance the reliability of SOC estimation under complex time varying conditions. First, a SOC estimation model of LIB based on Thevenin model was established. Then, a spatiotemporal neural network (NN) modeling method was developed to identify the unknown model parameters. In this modeling process, time NN and space NN models were constructed to describe the nonlinear dynamic of battery parameters at SOC scale and temperature scale, which improved the parameter identification accuracy under time-varying ambient temperature conditions. Finally, the spatiotemporal NN model was combined with ECM, and UKF algorithm was used to conduct SOC estimation. LIBs were tested under dynamic load and time-varying temperature conditions, and laboratory scale experimental results show that the proposed method has excellent accuracy and reliability for estimating battery SOC under time-invariant and time-varying temperature conditions.