The growing applications of lithium-ion batteries worldwide demonstrate the importance of accurate estimation of the state of charge (SOC) of batteries, as this directly impact charging quality and battery performance, and serves as a key indicator for users. Despite the importance of SOC, there is no physical device to measure it from the battery terminal. This paper proposes a comprehensive nonlinear fractional order model incorporating key factors affecting SOC estimation such as battery hysteresis effect, temperature, and age of battery into the model structure. In order to acquire thorough and broad range of experimental data, a battery tester and a data logger were set up through gathering the data under different conditions over a substantial period of time. The electrical elements of the equivalent circuit model were obtained using a PSO algorithm, and then input into the state-space equations. Based on the model outputs, a look-up table was generated for various conditions. Being reasonably accurate, the table could reduce intensive calculations considerably. Proposed nonlinear fractional order model, significantly reduces the computational overhead compared to adaptive models. Therefor by considering the electrical parameters as constant values and transferring all non-linearity and changes such as hysteresis effect, temperature and life cycle of studied battery in VOC function, the experimental results confirm the accuracy of proposed model.

Keywords