Diagnosis of Lithium Ion Batteries for Power Electronic Applications

Abstract

This poster presents a lithium ion battery State-of-charge (SOC) and the capacity estimation technique for electric vehicle applications. The proposed strategy achieves online parameter estimation by making use of adaptive control theory and Kalman observer. The convergence and stability are guaranteed by Lyapunov’s direct method. The estimation strategy takes into account the surface temperature variation. Experiments have been carried-out under different operating temperature conditions and various discharge currents. The effectiveness of the proposed method is verified by experiments for different aging states.State-of-Health (SOH) is an important aspect in Battery Management Systems (BMS) since it is considered, as the battery’s energy. Therefore, bad SOH estimation ultimately results in damaging the battery and reducing its lifespan. Similar to other chemical-based energy storage systems, the battery’s use generates irreversible physical and chemical changes and hence, its performance tends to deteriorate gradually over its lifetime. Several studies have been presented for lithium-ion battery calendar aging and show an internal resistance increase and a capacity decrease. So, the definition of the End-of-Life (EoL) of the battery depends on these aging indicators. The limit is generally set to 80% of the nominal capacity.In our case, the cells are prepared at different State-of-Charge (SOC) values and stressed with different temperatures. The cell parameters as the internal resistance and the capacity are measured periodically with well-defined discharge conditions. In this study, a 20Ah, 3.2V Lithium iron phosphate battery, LiFePO4, is used in the experiments. The battery is characterized, initially and after each aging phase, at different constant currents and at different operating temperatures.In order to study the evolution of the OCV-SOC characterization during the battery lifetime, we have compared the OCV-SOC characterization for different aging states. The results shown that the different between the OCV-SOC estimation and measured is negligible from one aging state to another.Using the experimental protocol results for a constant current, the internal resistance and the capacity evolution for different currents and the different operating temperatures are measured and estimated, and compared.Conclusion : In this study, an online SoC and SOH estimation method is presented for lithium-ion batteries. The proposed strategy capitalizes on the capabilities of Kalman filtering for the design of an extended Kalman observer to estimate SOC. Moreover, the adaptive estimation technique achieves online robust SoH estimation. Unlike other methods, this paper presents an online diagnosis method considering the surface temperature variation of the battery. Moreover, only voltage, current and temperature measurements are required, which reduces the number of sensors compared to other methods. The effectiveness of the proposed online observer is shown through a set of experiments. Results highlight its good performance in parameters estimation for different conditions of operating temperature, discharging current and battery lifetime.