-
Augusto de Holanda B. M. Tavares
Electrical Engineering Department, Federal University of Campina Grande
-
Florian vor dem Esche
Institute of Flight System Dynamics, Technische Universität München
-
Saulo O. D. Luiz
Electrical Engineering Department, Federal University of Campina Grande
-
Florian Holzapfel
Institute of Flight System Dynamics, Technische Universität München
-
Antonio M. N. Lima
Electrical Engineering Department, Federal University of Campina Grande
Keywords:
Lithium-Polymer batteries, Nonlinear system identification, Optimization, Nonlinear modelling, Control oriented models
Abstract
This paper proposes a model for Lithium-Polymer batteries with diffusion dynamics through a lumped parameter approach in conjunction with an electrical model. A comparison of the performance of several optimization approaches to identify the model parameters from experimental tests is also conducted. Constant and pulsed current discharge experiments were performed on a set of three battery cells using a programmable DC load. The resulting data sets were used to obtain several model parameters using different optimization approaches. The predicted outputs of a discrete battery model using the parameters estimated by each algorithm were compared against the experimental data. The resulting models had an overall good performance, proving that the chosen modelling approach is applicable to Lithium-Polymer batteries, and that the choice of algorithm to identify the system parameters must be made with care.