Thermal and electrical performance assessments of lithium-ion battery modules for an electric vehicle under actual drive cycles

• Published in: Electric power systems research 2018-10, Vol.163, p.18-27
• Main Authors: Panchal, S., Mathew, M., Dincer, I., Agelin-Chaab, M., Fraser, R., Fowler, M.
• Format: Article
• Language: English
• Subjects: Artificial neural networks; Deceleration; Drive cycle; Driving; Electric cells; Electric potential; Electric vehicle; Electric vehicles; Lithium; Lithium-ion batteries; Lithium-ion battery; Modules; Neural network; Neural networks; Rechargeable batteries; State of charge; Temperature; Temperature distribution; Temperature profiles
• ISSN: 0378-7796; 1873-2046
• DOI: 10.1016/j.epsr.2018.05.020

•Installation of data logger in EV and data collection under real world drive cycles at different ambient temperatures.•Design and development of an experimental set-up for battery and drive cycle testing.•Battery thermal model development using ANN and validation with the road test data and development of mathematical functions.•Performance assessment and evaluations of various drive cycles into test bench.•Validation in terms of SOC, voltage, and battery module temperature for different drive. In this paper, both thermal and electrical performance evaluations of a lithium-ion battery pack using real world drive cycles from an electric vehicle (EV) are presented. For the experimental measurements, a data logger is installed in the EV, and the real world drive cycles are collected. The EV has three lithium-ion battery packs consisting of a total of 20 battery modules in series. Each module contains six series×49 parallel IFR 18650 cylindrical valence cells. The reported drive cycles consist of different modes: acceleration, constant speed, and deceleration in both highway and city driving at 2°C, 10°C and 17°C ambient temperatures with all accessories on. Later, the same drive cycles are conducted in an experimental facility where four cylindrical lithium-ion cells are connected in series, and both electrical and thermal performances are evaluated. In addition, the battery model is developed using artificial neural network, which is validated with the real world drive cycles. The validation is carried out in terms of voltage, state of charge (SOC), and temperature profiles for all the collected drive cycles. The present model closely estimates the profiles observed in the experimental data. Moreover, with this study, the mathematical function for the average temperature, SOC, and voltage prediction are developed with weights and bias values.