Online Parameter Estimation of an Electric Vehicle Lithium-Ion Battery Using AFFRLS

Abstract

The most commonly adopted techniques used to estimate the state of charge SOC of a battery rely on equivalent circuit model ECM, the problem is that battery equivalent model parameters vary with many factors such as SOC, temperature, battery aging and so forth, which lead to SOC estimation error. Therefore, it is critical to accurately identify these parameters. One technique, known as online parameter identification, in which parameters of the battery model are constantly updated can be implemented to solve this issue effectively. In this paper, we suggest a new algorithm AFFRLS (adaptive forgetting factor recursive least squares) to extract the parameter of the battery model, then to predict the output voltage, and compare it to the original RLS (recursive least squares). To assess these techniques, we used experimental data performed on the LG 18650HG2 Li-ion Battery. We supplied the data to the algorithms and compared the estimated output voltage for one dynamic profile named the urban dynamometer driving schedule UDDS. Results show that AFFRLS has low distribution in high error range up to 6.4% less than RLS, this means that AFFRLS has a better parameter identification. Keywords— adaptive forgetting factor recursive least squares (AFFRLS), lithium-ion battery, urban dynamometer driving schedule UDDS.