Modeling and Experimental Evaluation of Lithium Iron Phosphate Battery

Abstract

The main objective of this paper is to present lithium iron phosphate battery modeling and experimental evaluation. The modeling of the battery was performed using the Thevenin equivalent circuit model with two RC branches and the nonlinear least squares method with the Levenberg-Marquardt optimization algorithm for parameter estimation. The modeling technique presents the most applicable and trivial solution to study and describe the behavior and performance of the battery. The experimental evaluation was performed on a 7,7 kWh battery pack installed at the Institute of Energy Technology, Faculty of Energy Technology, University of Maribor. Based on the estimated parameters of the Thevenin equivalent circuit, the model results show good agreement with measurements. The average mean absolute error method was used for determining the accuracy of the estimated parameters, with the state of charge and voltage being 8,38 % and 0,34 %, respectively.