The objective of this thesis is to explore a switching-based approach to estimate the state of charge (SOC) of Li-ion batteries. The knowledge of SOC can be utilized to significantly enhance battery performance and longevity. The thesis first presents a brief discussion on various SOC estimation methods, such as coulomb counting, use of electrochemical model combined with Kalman Filtering and open-circuit voltage (OCV). Subsequently, emphasis is placed on the OCV-based method. The advantage of the OCV method lies in its simplicity. It obviates the need for modeling and lowers computational burden compared to model-based approaches. The method yields accurate SOC estimates if a long period of battery resting time (switch-off time) is allowed. For smaller switch-off durations, the accuracy of SOC estimation reduces. However, experiments show that Li-ion batteries could give acceptable SOC estimates due to their fast transient response during switch-off. In traditional usage scenarios, a switch-off interval may not be practical. However, in distributed power systems with multiple storage elements, a switch-off interval could be provided. Experiments are conducted to characterize the estimation error versus the switch-off time. To reduce the switch-off time to 30 second switch-off time and to increase the accuracy of SOC estimation, a method is proposed to extrapolate the OCV at infinite time from the measured OCV using a time constant. This leads to predicted OCV for infinite switch-off intervals. Experiments are conducted to confirm the improved SOC estimation using the proposed method. For experimentation, a commercially available LiFeMgPO4 battery module as well as a single cell LiFePO4 battery, is used.
Library of Congress Subject Headings
Lithium cells; Energy conservation
Department, Program, or Center
Mechanical Engineering (KGCOE)
Su, Yingchen, "Switching-based state-of-charge estimation of lithium-ion batteries" (2011). Thesis. Rochester Institute of Technology. Accessed from
RIT – Main Campus