Research on the Torque Density Optimization of a Semi-Embedded Permanent Magnet Wind Turbine Based on the Non-Dominated Sorting Genetic Algorithm II and Magnetic Pole Offset

To improve the torque density (TD) of a permanent magnet wind turbine (PMWT), this paper proposes a magnetic pole offset semi-embedded structure based on the traditional semi-embedded structure. Firstly, the principle of how magnetic pole offset can increase the torque is explained. Then, based on the non-dominated sorting genetic algorithm II (NSGA-II), the ratio k of the inner and outer diameters of the stator is optimized to make the motor quality and efficiency reach the best state. On this basis, the TD is further optimized by utilizing the magnetic pole offset angle. The results show that when the magnetic pole offset angle is 0.5°, the TD reaches the maximum value of 13.95 Nm/kg, with an increase of 3.33%. Finally, the no-load performance and load performance of the two structures are compared to highlight the advantages of the magnetic pole offset structure.