A Novel Capacity Estimation Model Based on Scaling Transform for Series-Parallel LiFePO4 Battery System

Abstract For battery energy storage systems with series‑parallel system (SPS), accurately estimating the capacity of each parallel pack in the SPS is crucial for improving the overall system efficiency and capacity utilization, and it serves as a foundational yet significant research topic for estimating system-level states (state of charge, state of energy, and state of health). In this study, based on the characteristic analysis of SPS, a novel capacity estimation model for SPS based on scaling transform is proposed. The model was verified through simulations of five different inconsistency configurations, and the average estimation error across all the simulation results was 0.49%. Two sets of experimental verification tests were performed, and the average error was approximately 1.77%. The results from both simulations and experiments demonstrate that the model is viable and achieves reasonable accuracy for the capacity estimation of battery packs within the SPS.