Lithium iron phosphate battery soc estimation research

Source: 2020 - 04 - 11 hits of this statement: time of lithium iron phosphate battery charged state estimation study of lithium iron phosphate battery charged state estimation is in order to better application of battery as a power battery. This article selects the second-order RC battery model, using the adaptive noise matching trackless kalman filtering method for battery charged state estimation, improve the accuracy of the kalman filtering algorithm. The results of simulation and experiment show that this algorithm has high estimation precision and good effect of soc estimation. The state of the battery charged calculation is the important precondition for BMS system. Accurate estimation of power battery charged state, can improve the safety performance of the battery, effectively protect battery and prolong the service life of the battery pack, raise the use efficiency of the battery. Power battery soc estimation of the difficulty lies in the complex dynamic characteristics of battery system. Therefore, the soc estimation is the key to establish proper battery model, choose appropriate estimate method. Commonly used battery model with electrochemical model, neural network model and equivalent circuit model. This article chose can accurately reflect the dynamic characteristics of battery, the second order RC equivalent circuit model. Kalman filtering algorithm can real-time tracking system of the state, is suitable for power battery charged state estimate. Kalman filtering algorithm is applied to the evaluation method of linear system, and the battery is a complex nonlinear system, so the use of Taylor expansion of linearization of nonlinear system extended kalman filtering ( EKF) Algorithm of EKF algorithm can be applied to a good battery soc estimation research, but the calculation process is relatively complex, the stability of the calculation is poorer, so this article use no trace kalman filter ( UKF) Algorithm, the algorithm based on UKF to UT transformation of system state variables, in line with the state variables can be converted into statistical properties of the state variables of a few sampling points, and then in the system equations for computing. UKF algorithm than the EKF algorithm is simple and stable. In order to further improve the calculation accuracy, using adaptive matching algorithm for real-time updates the system state noise and observation noise, can further improve the accuracy of the system equations and the accuracy of the algorithm.