Loading...
Development of physics-based estimation and prediction algorithms for real-time Li-ion battery control
Evans, Tyler A. P.
Evans, Tyler A. P.
Citations
Altmetric:
Advisor
Editor
Date
Date Issued
2025
Date Submitted
Keywords
Collections
Files
Research Projects
Organizational Units
Journal Issue
Embargo Expires
Abstract
The enclosed thesis develops estimation and prediction algorithms suitable for real-time control of lithium-ion batteries. Battery states that are closely tied to degradation are typically unmeasurable. Thus, important battery degradation behaviors must be controlled by a battery management system without direct sensor measurement. Physics-based electrochemical battery models are used to simulate battery dynamics. These physics-based models are too computationally expensive to perform real-time prediction. Thus, reduced-order models are developed to capture the desired dynamics related to battery degradation. Although batteries are the main focus of this research, the developed estimator-predictor algorithm is written generically and is applicable to any nonlinear system.
The thesis is generally formatted to (1) introduce general formulation of battery dynamics and the degradation modes of concern, (2) compare various methods of reduced-order model extraction, (3) extend the existing linear impulse-response estimator-predictor to nonlinear systems, and (4) validate estimator-predictor performance through a series of actuation profiles. The estimator derived in this work is based on the impulse-response model. Estimation is formulated as a constrained least-squares optimization problem. Linearization of the constraint allows the optimization problem to be solved analytically. Provided the estimation results, an additional optimization problem predicts an optimal actuation sequence. The overall control structure uses model-predictive control framework. This work concludes with a model-based simulation demonstrating the rapid charge of a battery in under 15 minutes while avoiding lithium plating, the leading mode of degradation during fast charging.
Associated Publications
Rights
Copyright of the original work is retained by the author.