Eventos Anais de eventos
COBEM 2023
27th International Congress of Mechanical Engineering
ELECTRIC VEHICLE BATTERY MODEL IDENTIFICATION FROM DATA COLLECTED IN DYNAMOMETER TEST CONDITIONS
Submission Author:
Joaquim Manoel Gonçalves , SC
Co-Authors:
Joaquim Manoel Gonçalves, Samuel Luna de Abreu, Murillo Stein, Sergej Diel, Amir Antonio Martins Oliveira, Rui Araújo
Presenter: Joaquim Manoel Gonçalves
doi://10.26678/ABCM.COBEM2023.COB2023-1629
Abstract
This study focused on developing a mathematical model for the State of Charge (SoC) of an electric vehicle battery. The research utilizes current, voltage, and temperature measurements obtained from a Nissan Leaf 2012 running on a chassis dynamometer at the Argonne National Laboratory (ANL). The SoC model is crucial in determining the available energy and range for the vehicle, and it plays a fundamental role in the design, simulation, and analysis of electric vehicle battery systems. The measurements were collected at a high sampling rate of 10Hz during a comprehensive two-and-a-half-hour test, which involved various drive schedules to ensure a complete battery discharge. The regenerative braking system in the vehicle generates charge and discharge currents that improved the quality of the data used in the development of the proposed methodology. It employed the least squares method to correlate the collected data and identify the model's parameters and coefficients. The developed SoC model consists of an equation that considers the battery voltage as a function of SoC, temperature, and current. It is divided into two parts: the pseudo-open circuit voltage (pseudo-OCV) curve and a complementary time series of currents with four different delay terms. Other variables such as power, energy, and internal resistance were analyzed. The identification of the battery's internal resistance allowed for the determination of heat losses and energy efficiency (98%). The errors in the developed model are uniformly distributed and the average of the absolute errors for the voltage curve is 0.25%, while considering the inverse solution process, the average of the absolute errors for the SoC is 2.5%.
Keywords
battery SoC model identification, li-ion batteries, Electric Vehicle Batteries
