Pack Level State-of-Power Prediction for Heterogeneous Cells

Accurate prediction of available power in battery packs is crucial for managing performance in automotive and grid storage applications. A battery pack is composed of many cells, which have inherent cell-to-cell variation. This not only complicates the power estimation problem, but also adds complexity in ensuring that all cells remain in a safe operating regime. This paper presents a methodology to estimate the state of power (SOP) of a battery pack, composed of series connected heterogeneous cells. The presented SOP framework combines an interval prediction algorithm, with a modified reference governor. The concept of interval prediction accounts for cell-to-cell variability by predicting bounds that enclose all the states of all the cells at any given time. The proposed algorithm accurately predicts pack power without fixating on any individual cell. This makes the presented methodology computationally efficient and scalable to any number of heterogeneous cells.