Abstract
Online prognostics of the battery capacity is a major challenge as ageing process is a complex phenomenon, hardly directly measurable. This paper offers a new methodology for real-time estimating of the global battery performances for Electric Vehicle (EV) use. The presented data-driven framework build a model based on the modifications in battery signals behavior, according to the performance level. A first pattern extraction step consists in the selection of battery signals corresponding to specific acceleration profiles in real uses, allowing to highlight the battery behavior. These extracted voltage and current patterns are then considered to determine the battery behavior for each State of Health (SOH) feature. Studied patterns are compared using signal processing techniques, allowing the estimation of the battery performance, through statistical learning methods. The application of signal processing and Relevance Vector Machines (RVM) modelwith multiple kernels, provides a powerful tool to diagnose battery health online, only based on real signals. Furthermore, this methodology also allows the prediction of battery Remaining Useful Life (RUL) during real use. The proposed algorithm is validated using datasets from real EV uses. Presented diagnostics results on real data demonstrate the good accuracy of this new framework for battery SOH prognostics in real-time constraints, with uncontrolled conditions.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
