Li-ion battery aging tests and data analytics

• Main Author: Tan, Zheng Nan
• Other Authors: Xu Yan
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2024
• Subjects: Engineering; Li-ion battery data analytics; Li-ion battery aging tests
• Online Access: https://hdl.handle.net/10356/181653
• Institution: Nanyang Technological University

This project focuses on predicting the discharge capacity of lithium-ion batteries using data-driven models, with specific attention to two health indicators: Time Interval of Equal Discharge Voltage Difference (TIEDVD) and Incremental Capacity Peak (ICpeak). The study investigates the efficacy of these indicators in tracking battery degradation over repeated charge-discharge cycles. The TIEDVD was used to measure Voltage drop over time, while ICpeak was derived from incremental capacity analysis, capturing shifts in battery behavior. Using a framework, a Gaussian Process Regression (GPR) model was developed using data from cells 1 to 4 for training and cells 5 to 8 for testing, to assess the generalizability of the model. Results demonstrated that both TIEDVD and ICpeak are strong predictors of battery degradation, as evidenced by high R-squared values and low error metrics (RMSE, MAE, and MAPE). The model accurately predicted the discharge capacity across test sets, showing good performance in unseen data. This project provides a foundation for further research into battery health monitoring and predictive modelling using machine learning and health indicators, with potential applications in extending the lifetime of battery systems.