Model Parameter Evaluation for Nickel-Manganese-Cobalt Cells: An Examination and Verification of Various Approaches

Lithium cells are typically modelled through equivalent electrical networks whose parameters are tuned to closely match measured data. In this way, complicated experimental procedures are often needed to correctly achieve the actual battery behavior. After the testing, simulation models are fitted on experimental data through optimization algorithms. The computational cost is high, and significant time can be required to obtain the parameters. This article analyzes and discusses these two aspects. First, it examines which tests are needed and whether their duration can be shortened. Second, it describes how to correctly perform parameter identification and explores whether generalizations can be made in their formulation. The accuracy of each proposed solution is detailed, including test durations and computational costs.