Yizhou Zhang, Electrical engineering

Yizhou Zhang is a PhD student in the research group Automatic Control, Division of Systems and Control

Discussion leader is Professor, Remus Teodorescu, Aalborg University

Examiner is Professor Torsten Wik, Division of Systems and Control

To receive the password for the online link please email phdadm.e2@chalmers.se 

Abstract

Lithium-ion (Li-ion) batteries play a pivotal role in transforming the transportation sector from heavily relying on fossil fuels to a low-carbon solution. But, as an electrochemical device, a battery will inevitably undergo irreversible degradation over time. Therefore, accurate and reliable aging diagnostics and prognostics become indispensable for safe and efficient battery usage. However, diverse aging mechanisms, stochastic usage patterns, and cell-to-cell variations impose significant challenges. As the importance of vehicle operating data is becoming more apparent, an increasing number of automotive companies are collecting battery field data.
In this thesis, a series of machine learning (ML) frameworks, using both field data collected during vehicle operation and laboratory cycling data, for battery aging diagnostics and prognostics is developed. Among these, a data-driven multi-model fusion method is proposed to accurately and robustly estimate battery capacity under real-world arbitrary usage profiles. Additionally, a battery aging prediction framework is developed based on a combination of offline global models created using different ML methods applying histogram operational data and cell individualized models that are online adapted. Finally, the thesis presents an early-life prediction pipeline leveraging time-series and histogram data, showing that these two feature sources are effectively interchangeable and complementary. These algorithms are extensively evaluated with various data sources of different battery kinds. The evaluation results indicate that the developed methods are accurate and robust, and more importantly, they are applicable to the harsh conditions encountered in real-world vehicle operations