부산대학교 도서관

Lithium-ion battery module temperature and ambient temperature are the most significant factors in analyzing the dynamic performance of the module and greatly influences the certainty of battery SOC estimation. SOC estimation plays an crucial role in the prediction of the remaining driving range of electric vehicles (EVs) and the optimal charge/ discharge status of the battery. The most popular and commonly used method for the estimation of SOC is based on its relationship with open-circuit voltage (OCV). However, this estimation results in errors due to the fact that both OCV and SOC are dependent on battery module temperature. To analyze this problem, an SOC estimation technique based on two temperature-based models integrated with OCV-SOC-temperature table has been presented in this paper. To estimate the SOC at different operating conditions, an auto-upgraded neural network model is developed. Three driving cycle tests, Indian Urban Driving Schedule (IUDS), Indian Highway Driving Schedule (IHDS) and Urban Dynamometer Driving Schedule (UDDS) are performed to test the batteries at different temperatures. The IUDS is used to identify the model parameters while IHDS and UDDS are used for the performance validation of the SOC estimation technique. This approach is efficient and authentic when battery module temperature is changing according to the loading condition and considering the cooling effect.