Temperature and state-of-charge estimation in ultracapacitors based on extended Kalman filter

The performance and life expectancy of ultracapacitors depend heavily on the operating voltage and temperature. In this paper, simultaneous estimation of state-of-charge (SOC) and temperature is achieved by applying extended Kalman filter (EKF) algorithm with only the terminal measurement of voltage and current. For the application of EKF algorithm, a nonlinear model which consists of a voltage-and-thermal-dependent equivalent circuit model and a thermal model is first developed. The parameters in the equivalent circuit model are identified by applying least squares method with weightings at different frequencies so as to achieve satisfactory prediction over the whole applicable frequency ranges. Experimental results demonstrate that the EKF-based estimator is crucial in providing accurate and consistent prediction of SOC and temperature in existence of modeling errors and measurement noises, especially during dynamic charge/discharge cycles at low temperature. The accurate estimation of SOC and temperature enables optimum energy and thermal management of ultracapacitors.

► Simultaneous estimation of ultracapacitor SOC and temperature is achieved via EKF algorithm.
► The extended model includes voltage-and-thermal-dependent equivalent circuit and thermal dynamics.
► Satisfactory prediction over applicable frequencies is achieved via weighted least squares.
► The impact of modeling errors becomes more significant at temperature below freezing point.
► The EKF-based estimator is robust to modeling errors and measurement noises even at low temperature.