Passive and active tracking of electrochemical impedance of a drone battery
Résumé
More and more applications rely on batteries, for instance in the field of transport (electrical vehicles), of smart grid, and connected objects. Therefore, the development of battery has become a crucial issue. Unfortunately, a battery is a complex electrochemical system which depends on many parameters and whose performance deteriorates over time. Thus the development of an efficient battery management system (BMS) to avoid damages, to extend lifespan and to optimize the use of batteries is a priority. One way to obtain interesting information representative of the present state of the battery is to estimate its electrochemical impedance. If most of impedance estimation methods rely on the hypothesis of linearity and time invariance, the developed methods intentionally move away from this hypothesis. The goal of the presented work is not only to accurately estimate the impedance over a predetermined frequency range thanks to an algorithm which can be easily implemented in an embedded system but also to be able to track the temporal variations of the impedance. Different methods to estimate impedance both in time and frequency domains are developed and compared. Indeed the aim is first to precisely follow the evolution of the battery impedance in both active and passive identification cases. The evolutionary impedance estimation is then applied to a drone battery. The impedance estimator is used to evaluate the state of charge of the drone battery and other crucial indicators like the remaining flight time thanks to an extended Kalman filter.
Origine : Fichiers produits par l'(les) auteur(s)
Loading...