Study of Remaining Useful Lifetime Prediction for Electrolytic Capacitors based on Accelerated Aging Tests

In this paper, the Remaining Useful Life (RUL) of the electrolytic capacitor is estimated through the amount of change in capacitance and ESR. RUL is predicted by the conventional model and the amount of change in capacitance and ESR according to the operating time. This paper supplements the shortcomings of the conventional model that calculates the endpoint of life according to the operating conditions of the system. Therefore, the purpose of this text is to calculate the endpoint considering the aging state and speed of the capacitor. In order to confirm the degree of aging over time, the accelerated aging experiment of capacitors was conducted under two conditions of temperature and voltage charge/discharge stress. And capacitor state parameters, Capacitance and ESR, were measured. After predicting the endpoint through the conventional model, the error with the actual measured experimental value is predicted through data fitting. As a result, both types of aging showed high accuracy when data with a large degree of aging was used. In addition, the higher the tendency of the error between the calculated value of the conventional model and the actual value of the experiment, the higher the estimation accuracy. Therefore, in this paper, when applied to the algorithm proposed for voltage charging and discharging aging rather than temperature aging, it was confirmed that the estimation performance was superior to that of the conventional model estimation value. Copyright © The Korean Institute of Electrical Engineers.