Open AccessDevelopment of a Neural Network Model for SoH of LiFePO4 Batteries under different Aging Conditions
Author(s) -
Tsun-Cheng Kuo,
K. Y. Lee,
MingHsi Chiang
Publication year - 2019
Publication title -
iop conference series. materials science and engineering
Language(s) - English
Resource type - Journals
eISSN - 1757-899X
pISSN - 1757-8981
DOI - 10.1088/1757-899x/486/1/012083
Subject(s) - battery (electricity) , fade , artificial neural network , voltage , electric vehicle , battery capacity , computer science , power (physics) , current (fluid) , automotive engineering , state of health , electrical engineering , engineering , artificial intelligence , quantum mechanics , operating system , physics
LiFePO 4 batteries have a variety of superior properties, such as higher power densities, higher capacities, longer lifetimes and better safety. For these reasons, LiFePO 4 batteries are used extensively in electric vehicles and energy storage devices. However, there is an issue with the battery capacity in that it begins to rapidly fade after a certain number of charge and discharge cycles under compound influence of temperature and discharging current, which may lead to safety concerns. Therefore, it is very important to investigate the characteristics (voltage, current and capacity) of LiFePO 4 batteries in relationship to the number of cycles and environmental temperature. In this paper, for the sake of high efficiency and safe operation of LiFePO 4 batteries, we propose a Back Propagation neural network (BPNN) model which estimates the state of health (SoH) of the battery, so that the accumulated error of the capacities under different operating environments can be corrected. The accuracy of the model was verified in an electric vehicle with an average error of only 1.56%. The results show that the proposed model is satisfactory.