Nondestructive Pulse Testing to Estimate a Subset of Physics-Based-Model Parameter Values for Lithium-Ion Cells

Battery-management systems require mathematical models of their cells to be able to compute estimates of state-of-charge, state-of-health, state-of-power or -function, and state-of-energy. These models may be empirical in nature, like equivalent-circuit models. Alternately, they may be based on physical principles, and we believe that such physics-based models provide benefits over empirical models in terms of their ability to be leveraged by battery-management-system algorithms to enhance battery safety, lifetime, and performance. A roadblock to using physics-based models is the difficulty in determining their parameter values to model a physical cell accurately and inexpensively. This paper proposes a method to identify a subset of the physics-based model parameter values that can be performed with the kind of equipment found in many battery labs and which does not require teardown of the cell. It applies a very short duration pulse to the cell to measure instantaneous resistance as a function of state of charge and amplitude. These resistances are then fit to a reduced-complexity physical model to determine model parameter values. The method is first applied to a virtual cell in simulation to explore its features and limitations, and is then applied to estimate parameter values for a commercial automotive lithium-ion cell.

Automated summary

Battery-management systems rely on mathematical models to estimate battery states. Physics-based models offer advantages over empirical models, but determining parameter values can be challenging. This paper proposes a method to identify a subset of physics-based model parameter values without the need for cell teardown, making it accessible for battery labs and applications.