Design Optimization of Multi-Layer Permanent Magnet Synchronous Machines for Electric Vehicle Applications

Koua Malick Cisse, Sami Hlioui, Mhamed Belhadi, Guillaume Mermaz Rollet, Mohamed Gabsi, Yuan Cheng
2021 Energies  
This paper presents a comparison between two design methodologies applied to permanent magnet synchronous machines for hybrid and electric vehicles (HEVs and EVs). Both methodologies are based on 2D finite element models and coupled to a genetic algorithm to optimize complex non-linear geometries such as multi-layer permanent magnet machines. To reduce the computation duration to evaluate Induced Voltage and Iron Losses for a given electrical machine configuration, a new methodology based on
more » ... metrical symmetries and magnetic symmetries are used and is detailed. Two electromagnetic models have been developed and used in the design stage. The first model was the stepped rotor position finite element analysis called abc model which considered the spatial harmonics without any approximation of the waveform of flux linkage inside the stator, and the second model was based on a fixed rotor position called dq model, with the approximation that the waveform of flux linkage inside the stator was sinuous. These two methodologies are applied to the design of a synchronous machine for HEVs and EVs applications. Design results and performances are analyzed, and the advantages and drawbacks of each methodology are presented. It was found that the dq model is at least 5 times faster than the abc model with high precision for both the torque and induce voltage evaluation in most cases. However, it is not the case for the iron losses computation. The iron loss model based on dq model is less accurate than the abc model with a relative deviation from the abc model greater than 70% at high control angle. The choice of the electromagnetic model during the optimization process will therefore influence the geometry and the performances of the obtained electrical machine after the optimization.
doi:10.3390/en14217116 fatcat:l6lggxaaujg2vigy4u36g22pry