Charger Integrated Coestimation of Parameters and States of Battery

Abstract

Accurate parameter and state estimations of batteries are crucial for increasing safety and reliability. The ageing of batteries leads to electrochemical changes, changing the impedance and the charge-discharge characteristics. The impedance and derived equivalent circuit model parameters values support state estimation algorithms. Hence, if the parameters of the battery are not updated at regular intervals of time or usage, the state estimation will be erroneous. The battery management system, which performs the state estimation, is limited in functionality and accuracy due to dependency on predefined parameter values fed during the initial set-up. Hence, this work proposes charger-side online parameters and state estimation algorithms based on the impedance and the equivalent circuit parameters determined during the start of charging. The accuracies of the proposed algorithms are verified experimentally for two batteries: a new and an old lithium iron phosphate battery. Further, the algorithms are tested for four types of charging: constant current, constant current-constant voltage, pulse charging without discharge, and pulse charging with discharge. The experimental results show the suitability of the proposed algorithms for estimating battery parameters and states for both batteries. Moreover, the proposed algorithms are suitable for other Li-ion battery chemistry also.