State Of Charge Estimation In Electric Vehicles At Various Ambient Temperatures

The battery state of charge estimation at various ambient temperatures is critical to keep the electric vehicles safety. To solve the problem of battery model parameters vary with temperature, this work proposes a lithium-ion battery model with temperature effect and a state of charge estimation method at various ambient temperatures. The battery capacity and the open circuit voltage are fitted to establish the connection with temperature, respectively. Dual extended Kalman filters estimate the battery impedance. The experiments show that the lithium-ion battery model with temperature effect has high accuracy at different temperatures. Low temperature has a significant impact on battery model parameters. The proposed method does not need to store many battery model parameters' offline data and reduces the amount of experimental calibration for battery model parameters. The proposed method has a maximum error within 2% at various ambient temperatures. Moreover, the proposed method is robust to the initial state of charge value.