Determination of the load capability for a lithium-ion battery pack using two time-scale filtering

Abstract

Accurate determination of the continuous and instantaneous load capability is important for safety, durability, and energy deployment of lithium-ion batteries. It is also a crucial challenge for the battery-management-system to determine the load capability of a pack due to inevitable differences among in-pack cells. To overcome these challenges, a flexible battery equivalent-circuit-model is first developed without prior knowledge of in-pack cells’ capacity and internal resistance. Then, a method for determining the load capability of a pack is proposed based on a two time-scale filter using a combination of recursive-least-square-method and Kalman-filter. Distinguished from traditional model-based state-of-charge estimation strategies, only the terminal voltages and currents are used to estimate the open-circuit-voltages of each cell. The temperature and hysteresis effects are compensated to improve estimation accuracy. To reduce the computational complexity, a two time-scale scheme is designed to monitor the state-of-charge and model parameters of in-pack cells, which are further used to determine the peak power of the battery pack. Experiments and simulations conducted on LiFePO4 battery pack are employed to verify the performance of the proposed approach under dynamic operating currents and temperatures. The results indicate that the proposed approach is suitable for determining the load capability of a battery pack.