Dual particle swarm optimization based data-driven state of health estimation method for lithium-ion battery

Abstract

Accurate estimation of Li-ion battery state of health (SOH) is essential to ensure battery safety and vehicle operation. Here, this paper proposes a dual particle swarm optimization algorithm-extreme gradient boosting algorithm (DP-X) with the battery's charging voltage and incremental capacity (IC) data. First, the features are extracted from the voltage curve and the IC curve of each charging cycle through curve compression and interpolation. Then, this paper utilizes the PSO-XGBoost (P-X) algorithm to optimize the selected features and reduce the dimensionality of the features. Finally, the P-X algorithm was applied to combine with the optimized features to adjust the model's hyperparameters and estimate the SOH. Experimental results show that the maximum SOH estimation error of the dual P-X algorithm is less than 2 %.