SoC Estimation in Lithium-Ion Batteries with Noisy Measurements and Absence of Excitation

Abstract

Accurate State-of-Charge estimation is crucial for applications that utilise lithium-ion batteries. In real-time scenarios, battery models tend to present significant uncertainty, making it desirable to jointly estimate both the State of Charge and relevant unknown model parameters. However, parameter estimation typically necessitates that the battery input signals induce a persistence of excitation property, a need which is often not met in practical operations. This document introduces a joint state of charge/parameter estimator that relaxes this stringent requirement. This estimator is based on the Generalized Parameter Estimation-Based Observer framework. To the best of the authors’ knowledge, this is the first time it has been applied in the context of lithium-ion batteries. Its advantages are demonstrated through simulations.