A hybrid algorithm based on beluga whale optimization-forgetting factor recursive least square and improved particle filter for the state of charge estimation of lithium-ion batteries

Battery state of charge ( SOC ) is crucial in power battery management systems for improving the efficiency of battery use and its safety performance. In this paper, we propose a forgotten factor recursive least squares (FFRLS) method based on the beluga whale optimization (BWO) and an improved particle filtering (PF) algorithm for estimating the SOC of lithium batteries with ternary lithium batteries as the research object. Firstly, to address the accuracy deficiencies of the FFRLS method, the optimal parameter initial value and the forgetting factor value are optimized by using the BWO algorithm. Secondly, the adaptive simulated annealing algorithm (ASA) is introduced into the particle swarm optimization (PSO) to solve the sub-poor problem of traditional particle filtering. Experimental validation is performed by constructing complex working conditions, and the results show that the maximum error of parameter identification using the BWO-FFFRLS algorithm is stable within 2%. The MAE and RMSE are limited to within 2% when the ASAPSO-PF algorithm is applied to estimate the SOC estimation under Beijing Bus Dynamic Stress Test (BBDST), Dynamic Stress Test (DST), and Hybrid Pulse Power Characterization Test (HPPC) working conditions, indicating that the proposed algorithm has strong tracking capability and robustness for lithium battery SOC .