Finite Control Set Model Predictive Control Based on In-Situ Junction Temperature for Reliability Enhancement of Power Converters

In power converters, power semiconductor devices are often the components most prone to failure. As such, monitoring their health is crucial for enhancing the reliability of power electronics systems. In particular, the junction temperature of a power MOSFET can be used as an indicator of the stress experienced by the device. Therefore, it is possible to use the junction temperature to determine if the converter is operating within an acceptable operational range. The reliability of power semiconductors can be significantly improved when junction temperatures are maintained at lower levels. Based on the Arrhenius equation it is commonly accepted that a device’s lifetime doubles for every 10°C temperature reduction. To achieve this, a conventional finite control set model predictive control (FCS-MPC) method using a secondary problem formulation to reduce power loss and mitigate thermal stress through a thermal model is introduced. Based on this foundation, this paper proposes an FCS-MPC method based on in-situ junction temperature to reduce power loss and thermal stress, thus enhancing reliability of the power converters. First, the junction temperature of silicon carbide (SiC) devices is obtained by measuring their on-state resistance through double pulse tests. Subsequently, the relationship among junction temperature, on-state resistance, and current is described as an $R_{dson}-T_{j}$ map. Finally, the in-situ junction temperature based FCS-MPC approach for power loss reduction and thermal stress mitigation has been presented. The effectiveness of the proposed solution has been verified through simulation results, demonstrating a 25% reduction in junction temperature compared to the traditional SPWM method.