Parameter identification and SOC estimation of lithium-ion batteries based on AGCOA-ASRCKF

The state of charge (SOC) is an important parameter in battery management systems (BMS), and its accuracy is very important. In this paper, a co-estimation with the adaptive global optimal guided coyote optimization algorithm and the adaptive square root cubature Kalman filter (AGCOA-ASRCKF) is used to perform the parameter identification and SOC estimation of a lithium-ion second-order RC model. The AGCOA effectively solves the problems where traditional heuristic algorithms tend to fall into local optimum and have a slow convergence speed. The AGCOA can accurately identify the parameters of the battery model. At the same time, when compared with the cubature Kalman filter, the ASRCKF introduces a square root filter and adds a residual sequence to adaptively update the covariance of the process noise and measurement noise, which improves the estimation accuracy of the SOC. The method proposed in this paper is verified by intermittent constant current test, dynamic stress test, and the federal urban driving schedule. Simulation results show that a high-precision battery model can be established by AGCOA-ASRCKF. In addition, the predicted value of the terminal voltage is basically consistent with the actual value. At the same time, the SOC estimation error can be controlled to within 1.5%, and the algorithm has good robustness and reliability in the presence of errors in the initial SOC.