This content is not included in your SAE MOBILUS subscription, or you are not logged in.
An Adaptive Neuro-Fuzzy Inference System (ANFIS) Based Model for the Temperature Prediction of Lithium-Ion Power Batteries
- Bin Fan - China Automotive Technology and Research Center Co., Ltd. (CATARC) ,
- Chunjing Lin - China Automotive Technology and Research Center Co., Ltd. (CATARC) ,
- Fang Wang - China Automotive Technology and Research Center Co., Ltd. (CATARC) ,
- Shiqiang Liu - China Automotive Technology and Research Center Co., Ltd. (CATARC) ,
- Lei Liu - China Automotive Technology and Research Center Co., Ltd. (CATARC) ,
- Sichuan Xu
ISSN: 1946-4614, e-ISSN: 1946-4622
Published August 14, 2018 by SAE International in United States
Citation: Fan, B., Lin, C., Wang, F., Liu, S. et al., "An Adaptive Neuro-Fuzzy Inference System (ANFIS) Based Model for the Temperature Prediction of Lithium-Ion Power Batteries," SAE Int. J. Passeng. Cars - Electron. Electr. Syst. 12(1):5-11, 2019, https://doi.org/10.4271/07-12-01-0001.
Li-ion batteries have been widely applied in the areas of personal electronic devices, stationary energy storage system and electric vehicles due to their high energy/power density, low self-discharge rate and long cycle life etc. For the better designs of both the battery cells and their thermal management systems, various numerical approaches have been proposed to investigate the thermal performance of power batteries. Without the requirement of detailed physical and thermal parameters of batteries, this article proposed a data-driven model using the adaptive neuro-fuzzy inference system (ANFIS) to predict the battery temperature with the inputs of ambient temperature, current and state of charge. Thermal response of a Li-ion battery module was experimentally evaluated under various conditions (i.e. ambient temperature of 0, 5, 10, 15 and 20 °C, and current rate of C/2, 1C and 2C) to acquire the necessary data sets for model development and validation. A Sugeno-type ANFIS model was tuned using the obtained data. The numbers of input membership functions (MFs) representing the three input parameters of this model are 1, 2, 3, respectively. The input and output MFs are Gaussian curve and linear types, respectively. The optimization method is a hybrid one which is a combination of the back-propagation and the least squares methods. Compared with the validating data, the ANFIS model was able to accurately predict the battery temperature under various operating conditions. With fewer sensors for data acquisition and less computation complexity, this method could be a possible tool for the online temperature prediction of power batteries in electric vehicle applications.