Robust State of Health Estimation for Lithium-ion Batteries using Machine Learning

Machine learning (ML) technologies have gained considerable attention for state of health (SOH) estimation of Lithium-ion (Li-ion) batteries due to their advantages in learning the behavior of non-linear systems. ML methods do not require the battery physical modeling processes but rather map external characteristics of the battery to the loss in capacity. However, from the application perspective, there are still some challenges that need to be addressed, including the impact of data noise and data size on the estimation performance, the failure of features under variable operation conditions, the dependency on big data, and the difficulty of implementing complex algorithms in low-cost microprocessors. To cope with these issues, a systematic ML-based Li-ion battery SOH estimation framework is developed in this Ph.D. project, which has strong robustness to data size, data noise, and degradation conditions.

This Ph.D. project aims to identify and optimize robust ML-based SOH estimation algorithms for Li-ion batteries. The methods to improve the performance (i.e., the robustness, accuracy, and data size dependence) of ML-based SOH estimation have been systematically studied. The main findings of the Ph.D. thesis are summarized as follows:
• Selection of ML-based SOH estimation methods.
• A new lifetime model for calendar aging of Li-ion batteries.
• FE represents a robust feature for SOH estimation.
• Data noise suppression improves the SOH estimation accuracy.
• ELM with bagging technology is suitable for SOH estimation when limited training data is available.