A novel endurance prediction method of series connected lithium-ion batteries based on the voltage change rate and iterative calculation.

Abstract High-power lithium-ion battery packs are widely used in large and medium-sized unmanned aerial vehicles and other fields, but there is a safety hazard problem with the application that needs to be solved. The generation mechanism and prevention measurement research is carried out on the battery management system for the unmanned aerial vehicles and the lithium-ion battery state monitoring. According to the group equivalent modeling demand of the battery packs, a new idea of compound equivalent circuit modeling is proposed and the model constructed to realize the accurate description of the working characteristics. In order to realize the high-precision state prediction, the improved unscented Kalman feedback correction mechanism is introduced, in which the simplified particle transforming is introduced and the voltage change rate is calculated to construct a new endurance prediction model. Considering the influence of the consistency difference between battery cells, a novel equilibrium state evaluation idea is applied, the calculation results of which are embedded in the equivalent modeling and iterative calculation to improve the prediction accuracy. The model parameters are identified by the Hybrid Pulse Power Characteristic test, in which the conclusion is that the mean value of the ohm internal resistance is 20.68 mΩ. The average internal resistance is 1.36 mΩ, and the mean capacitance value is 47747.9F. The state of charge prediction error is less than 2%, which provides a feasible way for the equivalent modeling, battery management system design and practical application of pack working lithium-ion batteries. Graphical abstract Image Highlights • A novel endurance prediction method is proposed for the aerial LIB pack. • The voltage change rate and iterative calculation are launched. • The improved equivalent circuit model construction method is carried out. • Equivalent model parameters are determined and identified. • The generation mechanism research is investigated by using improved UKF algorithm. [ABSTRACT FROM AUTHOR]