A voltage dynamics model of lithium-ion battery for state-of-charge estimation by proportional-integral observer

Abstract

Observer-based algorithms for the state-of-charge estimation require setting up a dynamics model in battery management system of lithium-ion battery, and it is more important for the model to characterize the battery dynamics as much as possible. This is particularly challenging for formulating several characteristics into one battery dynamics model, for which the electrochemical reaction principle consists of some different disciplinary theories. In this article, a voltage dynamics model is developed for the design of a proportional-integral observer, which fuses the battery voltage characteristics, the current integral principle and the equivalent circuit model into a set of differential equations. Simultaneously, the model parameters are also updated in real time. Here we discuss a series of studies on the model-based algorithms that, collectively, develop a proportional-integral observer of how the state-of-charge is estimated by the use of the voltage dynamics model. Then, the proportional-integral observer algorithm is downloaded into a battery management system, which is deployed in a battery electric vehicle for the validation. Among the proportional-integral observer, the current integral principle and the extended Kalman filter, some comparative experiments are done with a same test cycle. Based on the experimental results and the statistical analyses, the proportional-integral observer by the voltage dynamics model is a good candidate for the state-of-charge estimation of lithium-ion battery in a real-world vehicle.