Automated Design of Traction Inverter to Achieve Optimal Performances Over the Whole Driving Cycle of an Electrical Vehicle

Inhalt:
The paper presents the automated design of a multilevel traction inverter for electric drive based on wide band gap semiconductors. Compared to other converters, the traction inverter operates constantly over an ultra wide range of voltage and current (speed and torque of the Electrical Machine). The driving cycles of the electrical vehicle are used as the selection criteria to sort several multilevel topologies in terms of efficiency. The automated design uses a multiobjective genetic algorithm to find the Pareto, i.e. the best trade-off for the converter design between power density, price and driving cycles efficiency. This approach makes it possible to compare the benefits from the wide band gap semiconductors like GaN or SiC introduction and associated multilevel topologies in the 800V system of an electrical vehicle.