Multi-Parameter Analysis of Li-ion Battery Degradation: Integrating Optical Fiber Sensing with Differential State of Health Metrics

Abstract

The reliability and safety of Lithium-ion batteries (LiBs) are of great concern in the energy storage industry. Nevertheless, the real-time monitoring of their degradation remains challenging due to limited quantitative metrics available during cycling. This study addresses this limitation by employing a novel approach that combines external optical fiber sensing with advanced data analysis techniques to comprehensively assess battery health. We engineered a non-invasive optical sensing platform using tandem pairs of polymeric and silica-based fiber Bragg grating (FBG) sensors affixed to the external surface of commercial Li-ion button cells, enabling simultaneous, real-time monitoring of device-level volume changes and thermal events over 600 cycles. Our analysis incorporated differential techniques to estimate the battery's state of health (SOH) based on capacity, strain, and temperature variations with respect to voltage. Additionally, we implemented and compared three deep learning models - Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU), and Artificial Neural Network (ANN) - to predict battery SOH over cycles. We were able to capture both continuous and spontaneous degradation events and provide unique insights into battery behavior across its lifecycle through differential analysis and new SOH metrics demonstrating high correlation with conventional measures. This multi-parameter approach, combining advanced sensing techniques with innovative data analysis and deep learning methods, contributes significantly to battery diagnostics, potentially improving reliability assessment, enhancing safety standards, and accelerating the development of more sustainable energy storage solutions.