Capacitor Capacitance Prediction Method Based on Time Series Method
Shenglei Wang, Ruishi Lin, Liang Bao, Ning Xue, Hao Li · 2024
The life prediction and degradation performance evaluation of capacitors are important means to analyze the reliability of electronic systems. Although the traditional life prediction methods based on physical models can explain the aging process of capacitors in theory, these methods often fail to accurately capture the complex characteristics and actual degradation laws of capacitors. Considering the significant influence of time factors and temperature environmental factors on capacitor aging, time series models have significant advantages in processing time series data and can effectively capture the trend and law of capacitance value changing over time. For this reason, the differential autoregressive moving average model (ARIMA) in time series is used to predict the capacitance value of capacitors. Experimental data of various types of capacitors are collected under four temperature environments: high temperature, low temperature, normal temperature and accelerated temperature. After data cleaning, the corresponding aging prediction model is constructed to predict the capacitance value of capacitors. The experimental results show that the capacitor aging prediction method based on ARIMA has good effect, high prediction accuracy, and can accurately reflect the aging process of capacitors under different temperature environments. Compared with traditional methods, the ARIMA model not only has a significant improvement in prediction accuracy, but also can provide more reliable aging evaluation in practical applications.