Based on the refinement successes to particle swarm optimization (PSO) methods, which include, namely, the introduction of an age variable, the proposal of new selection strategies to find the best solutions of the particle as well as for its neighbors, the design of a novel formula for velocity updating, the incorporation of an intensification search phase, and so on, an improved PSO method is presented. The experimental results reported indicate that the refined pinpointing search ability and the global search ability of the proposed method are significantly improved when compared to those of conventional PSOs

Quantum particle swarm optimization (QPSO), inspired from the basic concept of PSO algorithm and quantum theory, is a stochastic searching algorithm. However, the algorithm may encounter a premature convergence when dealing with multimodal and complex inverse problems. Thus, some improvements are introduced. More especially, one will randomly select the best particle to take part in the current search domain. Also, a mutation strategy is added to the mean best position, and an enhancement factor (EF) is incorporated to enhance the global search capability to find the global optimum solution and to avoid premature convergence. Moreover, some parameter updating strategy is proposed to tradeoff the exploration and exploitation searches. Experiments have been conducted on well-known multimodal functions and an inverse problem. The numerical results showcase the merit and efficiency of the proposed modified quantum inspired particle swarm optimizer (MQPSO).

This paper presents a stabilized nodal discontinuous Galerkin pseudospectral time domain (DG-PSTD) algorithm for fully anisotropic electromagnetic waves. This solver permits arbitrary high-order basis functions and adaptive hexahedral elements, thus very efficient for large-scale wave propagation in complex media. Maxwell’s equations are reformulated in a unified hyperbolic form, where a localized anisotropic Riemann solver is derived to serve as a numerical flux to exchange information across adjacent elements in the DG-PSTD scheme. This local analysis method also helps impose the time-domain anisotropic plane wave incidence in the total/scattering field framework. Numerical validations and applications demonstrate the efficiency, accuracy, and capability of this new high-order solver for 3-D large-scale generally anisotropic electromagnetic media.

Physics-informed neural network (PINN) has shown great potential in the inverse and parametric designing problems in electrical engineering. Moreover, most existing works on PINN are dedicated to computational fluids, and very few attentions have been paid to static and low-frequency electromagnetic near fields with multiple media in electrical engineering applications. In this work, a PINN for solving two-dimensional (2D) magnetostatic fields in electromagnetic devices and systems is proposed. The magnetic field intensity and the magnetic vector potential are solved by training a neural network which encodes partial differential equations and boundary conditions as residuals. The computation of the spatial derivatives of media constitutive parameters, which negatively impacts the training of PINN, is eliminated. A mesh-assisted non-uniform sampling method for the selection of collocation points is proposed to further improve the performance of PINN. The proposed PINN is verified by comparing its results with those of the finite element method in two 2D magnetostatic case studies. It is expected that this work will promote further applications of PINN in the modelling, numerical analysis, and parametric design of electromagnetic devices and systems.

This paper presents a novel magnetic-geared permanent magnet (MGPM) machine with dual-layer PM excitations, referred as dual-layer MGPM (DL-MGPM) machine. The key is to employ both the outer rotor and the modulation ring with PM-iron sequences, in which the PMs are separated by iron segments one by one, and can provide PM excitation and flux modulation simultaneously. Both the outer rotor and the modulation ring can achieve flux modulating, namely, bidirectional flux modulating, which can couple the magnetic fields excited by the two sets of PMs and armature currents effectively. The proposed DL-MGPM machine, as well as its existing counterpart which employing all the PMs on the outer rotor and referred as single-layer MGPM (SL-MGPM) machine, are optimal designed using finite element method coupled with genetic algorithm. The electromagnetic performances of the DL-MGPM machine and the SL-MGPM machine are investigated in detail and quantitatively compared. The results show that the proposed DL-MGPM machine can achieve higher torque density than the SL-MGPM machine, and is more suitable for low-speed high-torque direct-drive applications.

In this contribution the authors investigate a modular structure of Wireless Power Transfer system specifically dedicated to a quasistatic or dynamic operation mode in an urban environment. The proposed architecture has been simulated by using a commercial software. Characteristics of the interoperability are analyzed when the modular transmitting and receiving coils are in proximity, and the architecture has proven to be characterized by excellent performances in terms of small interaction between each single module. Finally, an improved coil design has been introduced which is characterized by good uniformity of the coupling coefficient when the receiver travels between two adjacent transmitting sections.