Modeling of Lithium-ion Battery Considering Temperature and Aging Uncertainties

Abstract

This dissertation provides a systematic methodology for analyzing and solving the temperature and aging uncertainties in Li-ion battery modeling and states estimation in the electric vehicle applications. This topic is motivated by the needs of enhancing the performance and adaptability of battery management systems. In particular, temperature and aging are the most crucial factors that influence battery performance, modeling, and control. First, the basic theoretical knowledge of Li-ion battery modeling and State of Charge (SoC) estimation are introduced. The thesis presents an equivalent circuit battery model based SoC estimation using Adaptive Extended Kalman Filter (AEKF) algorithm to solve the initial SoC problem and provide good estimation result. Second, the thesis focuses on the understanding of the temperature-dependent performance of Li-ion battery. The temperature influence is investigated through Electrochemical Impedance Spectroscopy (EIS) tests to enhance the theoretical basis understanding and to derive model compensation functions for better model adaptability at different temperatures. Third, the battery aging mechanisms are revisited first and then a series of aging tests are conducted to understand the degradation path of Lithium-ion battery. Moreover, the incremental capacity analysis (ICA) based State of Health (SoH) estimation method xiv are applied to track battery aging level and develop the bias correction modeling method for aged battery. In the final phase, the study of parallel-connected battery packs is presented. The inconsistency problem due to different battery aging levels and its influence to parallel-connected packs are discussed. Based on simulation and experimental test results, it shows that the current difference in parallel connected cells is increased significantly at low SoC, despite the battery aging levels and the number of cells in parallel. In total, this dissertation utilizes physics-based battery modeling and states estimation method to optimize battery management under temperature and aging uncertainties in electric vehicle applications. The unique contributions include developing analytical compensation functions to improve equivalent circuit battery model adaptability under temperature uncertainty and developing ICA based SoH estimation and battery modeling method to overcome aging uncertainty.