ONLINE STATE OF CHARGE AND STATE OF HEALTH ESTIMATION FOR LITHIUM-ION BATTERY BY THE NEURAL NETWORK MODEL

• Main Authors: Yu-chieh Chen, 陳育傑
• Other Authors: Chung-chun Kung
• Format: Others
• Language: en_US
• Published: 2019
• Online Access: http://ndltd.ncl.edu.tw/handle/p6cx36

碩士 === 大同大學 === 電機工程學系(所) === 107 === The online estimation of the state of charge and the state of health of lithium-ion battery (Li-ion battery) is studied in this thesis. The state of health (SOH) and state of charge (SOC) of Li-ion battery are two important items in Li-ion battery management systems. It is known that the usable capacity of a Li-ion battery varies with the SOH and the ambient temperature. In order to obtain good estimation of the SOC, it is necessary to have good estimation of the SOH of the Li-ion battery. A first-order RC equivalent circuit model (ECM) is adopted as dynamic model of the Li-ion battery. We design a characteristic test to find the relationship between the open-circuit voltage (Voc) and the SOC of the battery; the relationship between the ECM parameters (Rs) and the SOC of the battery under different SOH and different ambient temperatures. A back-propagation neural network is applied to estimate the SOH of the battery. The value of the Rs obtained from the characteristic test, SOC and ambient temperature (T) are adopted as the input data of the back-propagation neural network, and the usable capacity as output data. Thus, the SOH can be estimated by the usable capacity. Then, an online parameter update method is employed to find the values of the parameters (Rp, Cp) of the ECM. Finally, the adaptive extended Kalman filter (AEKF) is applied to estimate the SOC. The experimental results show the modeling error that caused by SOH and temperature effect can be compensated, and the SOC of the Li-ion battery can be accurately estimated by the proposed method.