A method to estimate battery SOH indicators based on vehicle operating data only. (15th June 2021)

• Record Type: Journal Article
• Authors: Vichard, L.; Ravey, A.; Venet, P.; Harel, F.; Pelissier, S.; Hissel, D.
• Is Part Of: Energy. Volume 225(2021)
• Publication Date: 2021-06-15
• Subjects: Battery SOH -- Battery modeling -- Vehicle operating data -- Electric vehicle; Power resources -- Periodicals; Power (Mechanics) -- Periodicals; Energy consumption -- Periodicals; 333.7905
• DOI: 10.1016/j.energy.2021.120235
• Languages: English
• ISSNs: 0360-5442

Abstract

Abstract: Batteries are multi-physical systems and during actual operating conditions they are submitted to variable ambient operating conditions which can affect the dynamic behavior and the degradation. Therefore, a good understanding of the dynamic behavior and the degradation laws under actual operating conditions is the key to a durability improvement and to the development of better energy management strategies. The purpose of the proposed study is to use an experimental database issued from a three years monitoring of a ten postal vehicle fleet to model the batteries with respect to operating conditions. Based on an electrical circuit model, an optimization algorithm and a Kalman filter, the scientific contribution is to propose a simple but efficient method, using vehicle operating data only, to estimate on-board the state of charge and state of health indicators linked to internal resistance and available capacity. The proposed model presents a very good accuracy and state of health indicators estimations show promising results. In the future, the proposed method could be applied on-board to estimate and analyze the state of health during the entire battery lifetime in order to provide an accurate state of charge estimation and to contribute to a better understanding of the degradation laws. Highlights: A database issued from the monitoring of ten postal hydrogen vehicles. A method to model battery directly from operating data considering thermal and SOC dependencies is given. Computing time is about few minutes. Normalized Root Mean Square error is lower than 0.02. A method to estimate Open Circuit Voltage, internal resistance and battery capacity directly online is given.