A single-point EIS measurement for SOC estimation of supercapacitor

A proper estimation of the state of charge (SOC) of an energy storage device plays a vital role in the efficient management of the battery system. In this research article, the SOC of a cylindrical hybrid supercapacitor (4000F, 4.2V) has been accurately estimated using a gaussian process regression (GPR) model and a post processing unit. The proposed technique is based on the electrochemical impedance spectroscopy (EIS) to train and test the GPR algorithm. The EIS measurements are collected at a single frequency (100 mHz) during the discharge cycle of the supercapacitor every 2% of the SOC while maintaining controlled environmental conditions (20°C, 50% humidity). In this way the model's training is independent from the current, voltage and temperature of the device under test. As a post processing mechanism, a linear three-point interpolation is performed followed by an error calibration method. The GPR model was able to estimate the SOC of the supercapacitor with a root mean square error (RMSE) of 1.56% with a standard deviation (SD) of 0.22. The post processing unit further improved the estimated SOC with a reduced RMSE of 0.92% and an SD of 0.18.