Predicting the Battery Residual Usable Energy under Dynamic Conditions: a Novel Adaptive Method with Enhanced Performance

Electric vehicle (EV) is a worldwide researching focus due to its environmental friendliness, but the inaccurate Remaining Driving Range (RDR) estimation hinders the EVs' popularity, and an accurate determination of the battery Residual Usable Energy (RUE) is the key factor to obtain a precise RDR value. A common RUE estimation method is based on State-of-Charge (SOC) estimation, in which the RUE is proportionally related to the current SOC. However, the battery voltage varies significantly under real-world conditions, and the traditional method results in certain estimation errors. An adaptive RUE prediction method (AEP) is introduced in this paper, in which the dynamic voltage is predicted based on the future discharge profile and a battery model, while the RUE is then calculated by the predicted voltage and current sequences. To prevent the model errors, the model parameters are onboard adapted by comparing the predicted and collected voltage values during discharge, through which the adaptive method could properly describe the dynamic discharge process. Based on a prismatic lithium Nickel-Cobalt-Manganese oxide/Graphite (NCM-G) battery, the RUE values are calculated through different methods under dynamic operation profiles with different discharge rates. As revealed from the results, the AEP could reduce the estimation error by more than 50% compared with the traditional method, and the accuracy is more satisfied for the cases with lower SOC limit. As a result, the presented adaptive method could effectively enhance the RUE accuracy and hence benefit the RDR prediction performance.