Open AccessSOC Estimation of Lithium Battery Based on Dual Adaptive Extended Kalman Filter
Author(s) -
Yongliang Zheng,
He Feng,
Wenliang Wang
Publication year - 2019
Publication title -
iop conference series. materials science and engineering
Language(s) - English
Resource type - Journals
eISSN - 1757-899X
pISSN - 1757-8981
DOI - 10.1088/1757-899x/677/3/032077
Subject(s) - extended kalman filter , kalman filter , control theory (sociology) , covariance , fast kalman filter , invariant extended kalman filter , recursive least squares filter , computer science , algorithm , covariance intersection , battery (electricity) , alpha beta filter , covariance matrix , adaptive filter , mathematics , artificial intelligence , statistics , moving horizon estimation , power (physics) , physics , control (management) , quantum mechanics
The estimation accuracy of single extended Kalman filter is not high, also it is affected by the initial value of state of charge (SOC). The second-order RC equivalent circuit model of lithium battery is established, and a joint algorithm, dual extended Kalman filter (DEKF) is proposed. Besides, the covariance matching theory is introduced for DEKF under the complex condition of uncertain noise statistical characteristics to improve the estimation accuracy. The improved DEKF is compared with another joint algorithms, i.e. recursive least squares and extended Kalman filter (RLS-EKF). Through the validation of battery test data, the modified dual extended Kalman filter based on covariance adaptive algorithm can realize real-time online estimation of battery SOC and time-varying parameters, and the estimation accuracy of lithium battery SOC and battery time-varying parameters is higher.