STATE-OF-CHARGE ESTIMATION FOR LI-ION BATTERIES USING SIGNAL PARAMETERS EXTRACTED FROM GUIDED ULTRASONIC WAVES

초록

This study aims to investigate the estimation of the state of charge (SOC) for lithium-ion batteries (LIBs) using signal parameters extracted in guided ultrasonic waves. Many efforts have been made to enhance the accuracy of the prediction of SOC in terms of algorithm structure and sensing methodology. Recently, it has been reported that ultrasonic wave signals could be correlated with SOC in LIBs, showing the potential of a new reliable method for SOC estimation. Hence, we studied SOC estimation using signal parameters extracted from guided ultrasonic waves measured on LIBs. The operation frequency was set to 170 kHz following the resonance frequency of the active sensor. The wave signals were collected at variable SOC under variable discharging rates of 0.5 to 0.9 C. The matching pursuit algorithm could capture parameters of measured wave signals, such as the extent of dispersion, amplitude, and asymmetricity. An artificial neural network model, trained with the estimated parameters and voltage, was capable of estimating the SOC of the LIB within a mean absolute error value of about 0.5%. The proposed methodology will provide a reliable SOC estimation for the safe usage of LIBs.

키워드