Hammerstein Model Based Battery SOC Estimation Considering Temperature Variation

Abstract

The State of Charge (SOC) of lithium-ion batteries is a crucial state in electric vehicle battery management systems. In this study, a Hammerstein SOC estimation model is constructed with three inputs (battery voltage, current, and temperature) and one output (battery SOC). Subsequently, by the key term separation principle, the key term (the output of the nonlinear part) is separated in the equation of the linear module, and is substituted with the equation of the nonlinear module. The model output is then represented as a linear autoregressive form with the minimum number of unknown parameters. Further, the adaptive moment estimation (Adam) algorithm is used to identify parameters of the Hammerstein SOC model. The simulation results demonstrate that the model can accurately and effectively estimate the SOC of lithium-ion batteries under varying road conditions and environmental temperature changes during automotive driving, and indicate that the Adam algorithm has the advantages of fast convergence and high accuracy compared to the batch gradient descent algorithm for model parameter identification.