Speeding Up Converter Design by Analytical Modeling and Genetic Optimization a More Electric Aircraft Example
Niklas Fritz, Rik W. De Doncker, Xuanlin Pan, The 10th International Conference on Power Electronics, Machines and Drives (PEMD 2020), 15-17 Dezember 2020.
A high gravimetric power density is a key requirement for power-electronic converters in next-generation more electric aircrafts. Analytical modeling of power-electronic converters including the weight of the components allows assessing the power density already in the design phase. However, more design variables exponentially increase computation time. This paper combines analytical modeling with genetic or evolutionary optimization to drastically reduce computation time. This way, even more design variables or a higher level of detail in modeling can be considered. The strategy is applied to a DC/DC converter interfacing the 270V DC bus in an aircraft with a supercapacitor-based active power filter. Both the single-objective genetic optimization of power density and the multi-objective genetic optimization of power density and efficiency are discussed. The results demonstrate the technological benefits of wide-bandgap power semiconductor devices in a more electric aircraft setting.