Influence of change in open circuit voltage on the state of charge estimation with an extended Kalman filter. (August 2017)

• Record Type: Journal Article
• Authors: Campestrini, Christian; Kosch, Stephan; Jossen, Andreas
• Is Part Of: Journal of energy storage. Volume 12(2017)
• Page Start: 149
• Page End: 156
• Publication Date: 2017-08
• Subjects: Ageing -- Filter tuning -- Battery modelling -- Validation -- Lithium-ion; Energy storage -- Periodicals; Energy storage -- Research -- Periodicals; 621.3126
• DOI: 10.1016/j.est.2017.04.011
• Languages: English
• ISSNs: 2352-152X

Abstract

Abstract : Highlights: Influence of ageing dependent open circuit voltage on the state estimation is shown. Two open circuit voltage determination methods and four temperatures are considered. Open circuit voltage dependency can be influenced by the Kalman filter tuning. Abstract: The Kalman filter is a commonly used state of charge estimation algorithm for lithium-ion cells. Therefore, the filter uses the correlation between the open circuit voltage and the state of charge as reference. The aforementioned relationship is largely affected by temperature and ageing. However, the effect of cell ageing on the filter accuracy is often neglected in literature. In this work we show the influence of an ageing dependent open circuit voltage on the state estimation with an extended Kalman filter. Therefore, the open circuit voltage (OCV) determined by constant current charge/discharge, and the OCV determined by charge/discharge steps, at four different temperatures in the range from 0 °C to 40 °C are considered. For each temperature, the open circuit voltage of cells with a state of health of 98.3%, 90.4% and 82.5% is changed during the validation process. Based on our measurements, we observed a root mean square error of approximately 3.5% in state of charge when the open circuit voltage of the new cell is compared to the aged cell. This value increases to a state of charge estimation error of up to 6.5% when the extended Kalman filter is applied. However, the open circuit voltage dependency on the estimation can be influenced by the filter tuning. In consequence, the open circuit voltage has to be adapted to the cell condition to fulfil the requirements of an accurate and reliable state estimation with open circuit voltage based algorithms.