This paper proposes a temperature reduction method using model predictive control with space vector modulation in SiC inverter-fed motor drive systems. Conventional approaches to temperature reduction using MPC often face limitations due to the tradeoff between temperature and power quality. The proposed temperature reduction method using MPC with space vector modulation (MPC-SVM) generates virtual voltage vectors to obtain high current quality. Furthermore, the long calculation time to calculate the cost function limits the practical implementation of the scheme. The limitation of calculation time is mitigated by suppressing the calculation of the cost function related to unnecessary voltage vectors. This article achieves effective temperature reduction by minimizing cost functions for current control and reducing loss. Power loss is predicted by constructing an energy-based loss model derived from the dwelling time of voltage vectors. The effectiveness of the proposed method is verified with simulation results.
