Battery Model Parameter Estimation Using a Layered Technique: An Example Using a Lithium Iron Phosphate Cell
By Robyn Jackey, Michael Saginaw, Pravesh Sanghvi, and Javier Gazzarri, MathWorks, and Tarun Huria and Massimo Ceraolo, Università di Pisa
Lithium battery cells are commonly modeled using an equivalent circuit with large lookup tables for each circuit element, allowing the model to closely match measured data. Pulse discharge curves and charge curves are collected experimentally to characterize battery performance at various operating points. Due to the number of values in the lookup tables, it can be difficult to fit the simulation model to the experimental data using optimization algorithms.
This challenge is addressed using a layered approach to break the parameter estimation problem into smaller tasks. The size of each estimation task is reduced to a subset of data and parameter values so that the optimizer can better focus on a specific problem. The layered approach was successful in fitting an equivalent circuit model to a lithium iron phosphate (LFP) cell data set to within a mean of 0.7mV residual error, and max of 9.2mV error at a transient.
Copyright © 2013 by The MathWorks, Inc. Published by SAE International, with permission.
This paper was presented at SAE World Congress.