An online method to simultaneously identify the parameters and estimate states for lithium ion batteries

Currently, state of charge (SOC) estimation on the basis of Kalman filter (KF) is realized to be applied online, but the parameters of the battery model that is implanted in KF are commonly identified offline. The offline identification is not only a time-consumed process but also provides the inaccurate results. Considering the complex and changeable operating conditions of batteries employed in electric vehicles, the parameters are also varied in fact and need to be identified online, by which the real state of the battery is reflected. In this study, the online identification of parameters and estimation of SOC are fulfilled simultaneously by using a novel algorithm named dual unscented Kalman filter (DUKF). Results show that the parameter identification on the basis of the DUKF accurately simulates the dynamic performance of the terminal voltage. By using the proposed algorithm and under three different operating conditions, the state of the battery is effectively estimated. The maximum error of SOC estimation is less than 3%, which is better than the results by using the extend Kalman filter (EKF) and unscented Kalman filter (UKF). Furthermore, the online identification of parameters that are changeable and related to the fading state of the battery, enables the state of health (SOH) to be estimated online. (c) 2018 Elsevier Ltd. All rights reserved.