Methods for predicting SOC of vehicle-mounted power battery of electric automobile

The invention discloses methods for predicting the SOC of a vehicle-mounted power battery of an electric automobile. The methods for predicting the SOC of the vehicle-mounted power battery of the electric automobile comprises the method for predicting the SOC in the charging mode, and the method for predicting the SOC in the discharging mode. The methods for predicting the SOC of the vehicle-mounted power battery of the electric automobile are based on an existing open-circuit voltage method and an existing ampere-hour integral method, an optimal battery model is obtained through curve fitting of the temperature, the voltage, the charging current and the discharging current of the battery, branch processing is conducted on the charging state and discharging state when an SOC value is estimated in real time, a corresponding SOC value is searched for in a model library by means of the immobilized battery model according to the temperature characteristic, the voltage characteristic and the current characteristic, of the battery, the SOC value is modified according to model errors, and the estimation accuracy is greatly improved.