Thermal-based finite control set model predictive control for IGBT power electronic converters

摘要

Thermal cycling is one of the main sources of aging and failure in power electronics. A possibility to reduce the thermal stress of semiconductors is to control the losses occurring in the semiconductor devices with the target to reduce the thermal cycles. This approach is known as active thermal control. The hardest limit of the existing active thermal control approaches is that they do not offer a general framework where the optimal switching sequence is selected in order to fulfill the applications demand and reduce the thermal stress in specific semiconductors. The goal is to achieve the minimum thermal stress for the best possible overall performance. For this purpose finite control-set model predictive control (FCS-MPC) seems the optimal approach because it allows including of non-linear thermal and lifetime related models into the control law. A precise control of the thermal stress in the semiconductors can be achieved as the optimal switching vector is directly applied to the physical system. This allows to avoid overrating of the used module or to increase its lifetime. In the paper the approach is proven using simulation and experimental results.