Accurate estimation of parameters and state of charge (SOC) is very important for the safe and reliable operation of solid-state batteries (SSBs). Traditionally, extended Kalman filters (EKFs) treat all data equally to estimate battery parameters and SOC. However, not all data are sensitive to estimation, and using insensitive data in estimation can lead to significant errors, especially in the presence of unknown interferences such as model uncertainty and measurement noise. In this study, two improved EKF methods are proposed to automatically identify sensitive data for estimation. First, the posterior Cramer-Rao lower bound (PCRLB) of the EKF is derived to indicate the performance of the EKF. Second, evaluation indexes based on the PCRLB are established to evaluate the sensitivity of EKF to data. Finally, two improved EKF methods with sensitive data selection mechanisms (DSMs) are proposed to estimate battery parameters and SOC, respectively. In addition, the entropy weight method is proposed to comprehensively evaluate the accuracy of parameter estimation. Simulation and experimental results demonstrate that the improved EKF methods significantly improve the parameter and SOC estimation accuracy of SBBs operating under different working conditions and temperatures.
Estimation of Parameters and State of Charge for Solid-State Batteries Based on Posterior Cramer-Rao Lower Bound Analysis
IEEE Transactions on Transportation Electrification ; 11 , 1 ; 3773-3784
2025-02-01
4356660 byte
Article (Journal)
Electronic Resource
English
British Library Online Contents | 2015
|AIAA | 2015
|