An improved variable forgetting factor recursive least square-double extend Kalman filtering based on global mean particle swarm optimization algorithm for collaborative state of energy and state of health estimation of lithium-ion batteries

Accurate assessment of SOE and SOH is a critical issue in the battery management system. This paper proposes an improved variable forgetting factor recursive least square-double extend Kalman filtering algorithm based on global mean particle swarm optimization to obtain a stable and accurate SOE and SOH at different aging levels and temperatures. Firstly, this paper establishes a framework for the parameter identification of variable forgetting factors recursive least squares algorithm based on the global mean particle swarm optimization. Then, proposing a global mean particle swarm optimization search mechanism centered on variable time double extended Kalman filtering. Finally, The proposed algorithm is validated on the hybrid pulse power characterization (HPPC) and Beijing bus dynamic stress test (BBDST) datasets. The experimental results show that the MAE and RMSE of the SOE results based on the HPPC condition are less than 0.0096 and 0.0153 at -5 °C and 15 °C. Similarly, the estimation results based on the BBDST condition are less than 0.0094 and 0.0102, respectively. The SOH estimation errors are less than 0.02. Therefore, the variable forgetting factor recursive least square-double extend Kalman filtering based on global mean particle swarm optimization algorithm can achieve accurate and stable SOE and SOH at different aging levels and temperatures.