Deep learning network based lifetime analysis of energy - fed traction power supply converter

This paper presents a life prediction method based on the parameters of actual operation history data collected by existing converter power unit sensors. Firstly, characteristics junction temperature curves forced air-cooled radiator and are extracted, deep learning neural network architecture is constructed characteristics. Then thermoelectric coupling model thermal resistance calculation theory established, cumulative loss obtained from measured data. The trained verified. Finally, distribution under different setting thresholds correlation analysis with obtained, which provides feasible scheme for parameter prediction.