An improved forgetting factor recursive least square and extended particle filtering algorithm for accurate lithium-ion battery state of energy estimation

State of energy (SOE) estimation of lithium-ion batteries is the basis for electric vehicle range prediction. To improve the estimation accuracy of SOE under complex dynamic operating conditions. In this paper, ternary lithium-ion batteries are used as the object of study and propose a hybrid approach that combines a particle swarm optimization-based forgetting factor recursive least squares method with an improved curve-increasing particle swarm optimization-extended particle filter algorithm for accurate estimation of the state of energy of lithium-ion batteries. Firstly, for the accuracy defects of the FFRLS method, the particle swarm optimization algorithm is used to optimize the initial value of the optimal parameters and the value of the forgetting factor. Secondly, the curve-increasing strategy is introduced into particle swarm optimization to solve the sub-poor problem of extended particle filtering. Experimental validation through different working conditions at multiple temperatures. The results show that the maximum error of parameter identification using the PSO-FFRLS algorithm is stabilized within 1.5%, and the SOE estimation error is within 1.5% for both BBDST and DST conditions at both temperatures. Therefore, the algorithm has high accuracy and robustness under different complex working conditions. The estimation results prove the effectiveness of the energy state estimation.