Overview of State of Power Prediction Methods for Lithium-ion Batteries

Abstract

With the large-scale application of lithium-ion batteries in smart grid and new energy vehicles, the accurate prediction of their charging and discharging capacity, namely peak power prediction, is very important to maintain the safe and reliable operation of the system. This paper analyzes the state of the art of state of power prediction methods for lithium-ion batteries from the single and system levels: ① For cell prediction methods, mainly including look-up table method, black box method, equivalent circuit model and electrochemical model method. The equivalent model method with multi-parameter constraint is emphatically introduced. The classification and comparative analysis of those methods are also carried out. ② For battery system, viewing from battery system model and state of power estimation methods, the state of power prediction algorithm of series and non-series battery system and the intelligent prediction method driven by big data are discussed. Moreover, the advantages and disadvantages of these methods and the application field are analyzed. ③ Combined with the development trends of next-generation cloud computing, big data, digital twin, etc., the state of power prediction methods of lithium-ionbatteries are forecasted, which provides some ideas for the development and application of battery all life cycle management technology.