Application-Driven Data Acquisition for Condition Monitoring of Power Semiconductors in Traction Inverter Applications

Abstract

Condition monitoring has proven to be an effective means of enhancing the reliability of power semiconductors. This article proposes an innovative and practical data acquisition scheme that can serve various condition monitoring purposes for power semiconductors in traction inverter applications. It is achieved through the dc-link discharging operation mode of the traction inverter after the system is shut down, minimizing the new risks introduced by condition monitoring into system operation, which is critical to the safety of traction applications. Moreover, due to the extremely low dynamics in the discharging operation mode, the sampling requirements for the health indicators and operating conditions of power semiconductors are greatly reduced, and the measurement disturbances from parasitic parameters are eliminated. The proposed method is promising for electric vehicle (EV) applications, as the required operation condition is an inherently programed and frequent mode. Compared with conventional online data acquisition schemes, it outperforms in complexity, accuracy, safety, and practicality. The proof-of-concept experiments based on a three-phase inverter prototype verify the effectiveness of the proposed method.