Determination of the entropic heat coefficient in a simple electro-thermal lithium-ion cell model with pulse relaxation measurements and least squares algorithm

Abstract

Simplified battery models provide a good trade-off between accuracy and efficiency and are crucial for the implementation in battery management systems for the prediction of the batteries’ electrical and thermal performance. The coupling between a thermal and an electrical model considers the heat generation and the temperature dependent behavior of the cell. However, the measurements for parametrization of such battery models are currently time consuming and complex. This work introduces a method for extracting the entropic heat coefficient as well as the electrical parameters from the same pulse relaxation measurements which saves plenty of time and effort. The pulse relaxation method is applied at four different temperatures and three different current rates to a LiCoO2 cell from LG Chem with 2.15 Ah. A subsequent least squares estimation between the measured and the simulated temperature profiles refines the extracted entropic coefficient. The comparison between the determined coefficients and the coefficients measured by the well-established procedure shows a good agreement. The simulated and measured surface temperature corresponds well and therefore validates the electro-thermal model with the determined parameters. The maximum temperature deviation between the simulation and the measurement takes 0.5 K at an ambient temperature of 15∘C and applied current pulses of I=1.29A.