Step-skew is a common technique for eliminating the cogging torque of a target harmonic order in permanent magnet synchronous motors (PMSMs). However, when step-skew is applied to the rotor, the cogging torque of the target harmonic order is not completely eliminated due to 3-D leakage flux. Therefore, the 3-D leakage flux should be considered in designing a PMSM with step-skew for cogging torque reduction. The most accurate way to consider the 3-D leakage flux is to perform 3-D finite element analysis (FEA), but it has the disadvantage of high computation time. To resolve this challenge, this paper proposes a design method that utilizes transfer learning to reduce the time for 3-D FEA while maintaining accuracy. Through the proposed method, a large amount of 2-D FEA-based data and a small amount of 3-D FEA-based data are used instead of a large amount of 3-D FEA-based data, with similar accuracy as using a large amount of 3-D FEA-based data, and the computational time is highly reduced. Finally, a prototype is fabricated and tested to verify the validity of the proposed design method for cogging torque reduction.
Linear compressors have a simpler mechanical structure and higher efficiency than the rotary compressors because they have lower frictional loss. Therefore, the linear oscillating actuator (LOA) is an attractive option for compressors owing to its high power density and efficiency. However, the complex structure of LOA such as segmented outer stator and mover leads to conduct 3-D finite element analysis (FEA) for calculating accurate iron loss, but it requires high computation cost. Thus, we proposes a method to calculate iron loss only using 2-D axisymmetric FEA considering the permeance in stator core. To compensate for the alterations in the magnetic flux density resulting from the structure of the mover and outer stator of the LOA, two equivalent coefficients are applied in permanent magnets in mover to correct the induced voltage and outer stator to correct the magnetic flux density. Through the utilization of these two equalization methods, iron loss can be accurately calculated using only 2-D axisymmetric FEA. The proposed method can be used to accurately determine the efficiency of the LOA without 3-D FEA, thus, making the LOA design process more efficient.
The dual inverter and dual stator system is suitable for air mobility due to its redundant system. Motor analysis for transportation should encompass not only electromagnetic characteristics but also drive and multiphysics systems to account for real operating conditions. To reduce the considerable time of system integrated analysis, especially those involving drive and multiphysics systems, a parameter-based model can be employed as a substitute for the finite element analysis (FEA) co-simulation. Therefore, an accurate and fast parameter-deriving method is required. The conventional method, the frozen-permeability method, is accurate but time-consuming because the dual stator system has doubled inputs. In this article, a parameter estimation method based on the incremental method is proposed to consider a dual stator motor. Subsequently, the frozen-permeability method and the proposed method are compared to assess their accuracy. The results show that the proposed method can acquire accurate data faster than the frozen-permeability method. In addition, experiments are conducted to validate the parameter-deriving methods.
2 July 2025 in IEEE Transactions on Industry Applications
This article proposes a method for estimating the 3-D FEA-based demagnetization ratio (DR), which serves as a key measure for assessing irreversible demagnetization, using deep transfer learning. The complex configuration of the LOA, such as the segmented outer stator and permanent magnets (PMs), reduces the accuracy of 2-D axisymmetric finite element analysis (FEA). While 3-D FEA provides a more precise DR estimation, its substantial computational cost poses a significant challenge. Therefore, we propose a deep transfer learning-based demagnetization analysis method that improves computational efficiency while preserving high accuracy. This approach takes into account the permeance in the stator core and circumferential leakage flux. By leveraging deep transfer learning, knowledge acquired from a large-scale 2-D axisymmetric FEA-based DR dataset is transferred to a limited 3-D FEA-based DR dataset, effectively enhancing deep neural network performance. The DR predicted by the proposed method was compared with the results obtained from 3-D FEA. Using the proposed method, the analysis time is significantly reduced compared to employing only 3-D FEA, while maintaining high accuracy. This demonstrates its potential for accurate and efficient demagnetization analysis, positioning it as a viable solution for LOA design optimization. The proposed method is validated through computer simulations and experiments under various demagnetization conditions.