To address the inadequate dynamic response characteristics of hybrid energy storage systems (HESS) in aviation dc microgrids under pulsed power loads caused by traditional PI control, this article proposes an optimized feedforward compensation strategy based on dynamic allocation of load-current and derivative weighting. This method bypasses the transient response limitations inherent in conventional PI control and incorporates a dynamic compensation term linked to the load current derivative directly into the duty cycle unit, enhancing the traditional dual-loop control framework. By perceiving real-time load current variations, the approach proactively adjusts power allocation coefficients among energy storage units. Compared with traditional PI control, it significantly reduces dc bus voltage recovery time under pulse load conditions, effectively enhancing bus voltage robustness and power quality. Furthermore, the article provides a theoretical analysis of the dynamic performance and stability of the proposed method from a system transfer function perspective. Finally, numerical simulations and experimental results validate the efficacy of the improved approach. The findings offer a theoretical foundation and technical reference for optimized power allocation design of HESS in high-reliability aviation dc microgrids.

Pseudo-Hermitian wireless power transfer (WPT) system relaxes the rigid limitations of the conventional parity-time (PT) WPT system, while its multimode characteristic needs to be clarified. To clarify the eigenmode selection mechanism, an accurate modeling of the gain element is usually required. Since it is unachievable to model all the existing gain elements accurately, a qualitative method would be more useful. Recently, the criterion of lowest gain is proposed, which claims that PT WPT and Pseudo-Hermitian WPT system will automatically stabilize at the eigenmode corresponding to a purely real eigenfrequency with the lowest gain. The criterion of lowest gain is a qualitative and empirical method without the requirement of accurate modeling. However, this article demonstrates that a stable eigenmode can still exist even if its corresponding gain is not the lowest and found that the selected eigenmode can be determined by the initial operating condition. Therefore, an extension of the gain criterion is proposed, which claims that for all the eigenmodes with a purely real eigenfrequency in the system, the eigenmode with the lowest gain is always stable. The higher the gain of other eigenmode is, the lower the possibility of its stability is. Finally, experimental results have confirmed the correctness of our work.

A new scheme for handling nonlinear terms is incorporated into the precise integration time domain (PITD) method for electromagnetic simulations in a nonlinear medium. Unlike the traditional scheme used in conductive nonlinear magnetic medium, the new scheme avoids matrix inversion, eliminating potential numerical instability caused by singular matrices and providing a more robust solution. Consequently, PITD no longer relies on the condition number of the coefficient matrix, successfully extending its applicability to Lorentz dispersive and Kerr nonlinear medium. Furthermore, as demonstrated by numerical examples, the new scheme performs better with large time steps than the traditional scheme, especially when the medium exhibits strong nonlinearity.

The interactions between electromagnetic (EM) waves and Kerr nonlinear media result in various appealing phenomena, including self-focusing, optical vortex soliton, etc. However, existing numerical methods for resolving Kerr nonlinearities demand multiple iterations or solutions to a huge number of cubic equations in each time advance, thus hardly achieving the optimal trade-off between accuracy and efficiency. In this article, an explicit non-iterative discontinuous Galerkin time-domain (DGTD) method is proposed for transient EM analysis involving Kerr media. The proposed method can be seamlessly integrated into the widely adopted Runge-Kutta-DGTD (RK-DGTD) method. The essence is the deliberate evaluation of the incremental electric field intensity, during which the only approximation required is to replace the current electric field intensity with that at the previous intermediate time instant of the RK scheme, so that implicit or iterative time stepping can be avoided. In addition, in regions with strong nonlinear effects or fine geometry, the computational burden is exacerbated by local fine meshes and the resultant tiny time-step size. Therefore, the local time-stepping (LTS) technique is further introduced to improve computational efficiency. Numerical results demonstrate that the proposed method can achieve high accuracy similar to those of Newton’s iterative method and direct method with the cubic formula, and consume roughly the same short execution time as that of the conventional explicit non-iterative method.

This article presents a modified analytical method for evaluating the apparent power in homogenized models of conductor arrays, with particular focus on structures exhibiting periodic and regular hexagonal symmetry. For analyzing the external effect, the extended Ollendorff formula is implemented. For evaluating the internal effect, by solving the diffusion equation with Bessel functions, the apparent power inside the conductor is obtained using the Poynting theorem. The total power of the unit cell is subsequently completed through the application of Amp & egrave;re's circuital law. The accuracy of the modified analytical solution is validated against detailed finite element simulations across a wide frequency range. Results show that the modified method significantly improves the prediction of reactive power, especially in the high-frequency regime, while maintaining excellent accuracy in active power estimation. The effectiveness of the method is further demonstrated in large-scale, homogenized domains composed of multiple periodic and hexagonal cells. Detailed field distribution comparisons confirm the validity of the homogenization process.

Powder cores are ideal materials for fabricating filter inductors in high-power power conversion systems (PCSs). Allowing these cores to operate in their soft saturation regions significantly reduces the volume, cost, and losses of the inductors; however, this introduces strong nonlinear current dependence, thereby invalidating traditional inductor design methods based on linear materials. This article proposes a comprehensive electrothermal co-optimization method for inductors that accounts for saturation effects, current ripple, losses, and thermal behavior. First, based on space vector pulsewidth modulation and an accurate reluctance model, a three-phase three-level PCS current ripple calculation method is established that accounts for saturation characteristics, providing theoretical support for filter design. Based on this, a core and winding loss model considering magnetic saturation was constructed. Second, to tackle the heat dissipation issues arising from high power density, an integrated thermal network incorporating conductive, convective, and radiative thermal resistance is constructed to accurately evaluate the performance of the inductor heat dissipation system. Finally, a parametric sweep optimization tool is implemented to support efficient design space exploration and performance evaluation. Comprehensive testing on a 215-kW/9-kHz PCS prototype verifies the reliability of the proposed method.

A discontinuous Galerkin time-domain (DGTD) solver is implemented for simulation of plasmonic nanostructures by coupling Maxwell equations with the generalized nonlocal optical response (GNOR) model. Unlike the Drude model and the nonlocal hydrodynamic Drude (NHD) model, GNOR captures both spatial-dispersion effects and diffusion induced size-dependent damping effects that become significant near the plasma frequency. The diffusion term in the GNOR model is treated implicitly to enhance numerical stability without compromising computational efficiency. Numerical results validate the DGTD solver and compare the Drude, NHD, and GNOR models, highlighting the effects of nonlocality and diffusion.

The current filament method (CFM) is a widely used and efficient approach for simulating coilgun systems, but its accuracy is highly sensitive to the precision of mutual inductance calculations. This article presents a novel method for calculating mutual inductance, in which the positions of the equivalent cells in the driving coil are determined by Gauss-Legendre quadrature points, replacing the conventional uniform cell distribution. This approach achieves a high convergence order, maintaining high accuracy with fewer cells. Furthermore, leveraging the structural and motion characteristics of the coilgun, a physics-driven search algorithm is developed to automatically determine the near-optimal number of Gaussian points for different coils, thereby eliminating the need for cumbersome manual adjustments. The proposed method and the search algorithm exhibit a strong synergistic effect: the higher error-convergence order of the proposed method significantly reduces the search domain and accelerates convergence compared with the original method.

The hybrid energy storage systems (HESSs), often configured with battery and supercapacitor (SC) combinations, can effectively regulate power imbalances between generation and loads within more electric aircraft (MEA) dc microgrids. However, traditional control exhibits limitation in precisely tracking high-frequency components within SC current loop. To address this deficiency, this article proposes a novel current control strategy enhancing SC current tracking capability in HESS. Departing from conventional approaches, the proposed strategy eliminates the proportional-integral (PI) controller from the SC current loop, instead integrating SC current regulation with pulse-width modulation (PWM) to form a modified PWM unit. Consequently, the input signal of this modified PWM unit derives from given function of the SC current, while the amplitude of the PWM carrier signal is reformulated to vary instantaneously with the SC current. The variable-carrier-based PWM method accelerates the realization of SC current synthesis. Furthermore, a small-signal model is also explored to deeply investigate the stability of the HESS. Subsequent simulations are conducted to verify the performance across multiple scenarios. Experimental validation conducted on a 25 kW HESS power stage demonstrates the theoretical feasibility of the proposed control methodology.

Based on the finite difference time domain (FDTD) method and the three-stage split-step precise integration time domain (SS2-PITD) method, a hybrid subgridding method is proposed for solving the multiscale problems. In the proposed method, the multiscale problem is decomposed into two types of sub-problems. The sub-problem of the extensive background region is simulated by FDTD with coarse grids. Considering the limitation of the FDTD Courant–Friedrichs–Lewy (CFL) condition, the sub-problem containing the small-scale fine structure is analyzed by SS2-PITD with fine grids to eliminate the small time step determined by the fine grid size. Hence, the time step sizes in the fine and coarse grid regions can be synchronized, so as to reduce the number of updating steps in the fine grid region and improve the computational speed. Numerical experiments are simulated to show the effectiveness of the proposed method.

Conventional numerical simulation methods may become challenging when handling complex geometries due to ineffective discretization and large number of degrees of freedom. Based on the integral representation of electric potential derived from the Poisson equation and Green’s theorem, a recursive estimate of the potential can be obtained, which can be realized by using the walk-on-spheres (WoS) algorithm. When abrupt change of material exists, the conventional WoS algorithm based on the mean value theorem becomes ineffective due to possible branches at the material interface, which is resolved by developing a probabilistic sampling strategy. The proposed method is mesh-free, progressive and highly parallelizable, showing great potential in handling Poisson boundary value problems with complex materials and geometries.

A novel multi-objective optimization approach for the magnetic design of power inductors is proposed, leveraging the deep reinforcement learning (DRL). The optimization problem is formulated using an artificial neural network (ANN) with the encoder-decoder architecture. The encoder processes the design specifications and material properties to ensure the model’s generalization, while the decoder indirectly generates the optimal design variables by modeling their probability distributions. Trained via reinforcement learning, the DRL-based method addresses the design task without iterations, significantly reducing computational time. Additionally, compared to the heuristic algorithms like non-dominated sorting genetic algorithms (NSGA-II) and the decomposition-based multiobjective evolutionary algorithms (MOEA/Ds), the proposed approach produces consistent Pareto fronts across multiple repeated runs and ensures a more uniform distribution of solutions. The effectiveness of the approach is further validated through a case study on a powder core inductor for inverters, highlighting its advantages in the convergence, stability, solution diversity, and efficiency.

Space-time modulated systems have recently emerged as a powerful platform for dynamic electromagnetic processing in both space and time. Most of the related research so far has assumed abrupt parameter profiles. This paper extends the field to generalized graded-index (GRIN) interfaces, which are both more practical than ideal profiles and offer new avenues for wave manipulations. It presents an exact solution for wave propagation across arbitrary space-time modulated GRIN interfaces and describes versatile chirping effects. The solution is based on a generalization of the impulse response method from linear time-invariant to linear space-time-varying systems. The proposed framework shows that space-time GRIN systems represent a novel approach for generating a new form of chirping that is not inherently based on dispersion, with promising applications in pulse shaping and signal processing.

When free electrons travel photonic crystals (PCs), multiple transition radiations occur, yielding highly directional resonance radiation if the emitted electromagnetic (EM) waves are coherent. In this work, we have investigated the resonance transition radiation regulated by the moiré lattices, which can offer greater flexibility in manipulating EM waves compared to conventional PCs, facilitating more intricate free-electron radiation phenomena. In particular, an exponential scattering matrix cascading approach is proposed, which can significantly decrease the necessary cascading steps, signifying a substantial reduction in computational resources. The angular energy density distribution is obtained and the velocity forbidden zone is observed. The radiation modes are categorized into s- and p-polarized types, determined by the inherent dispersion relation and iso-frequency contours. It is also found that the lattice band gap leads to a significant suppression of the radiation. Our work could provide a foundation for exploring the promising wave behavior within the moiré lattices.

A novel homogenization framework is proposed in this study for modeling multiscale Litz wire coils. The proposed method integrates the finite element method (FEM) with the extended Ollendorff formula to evaluate the magnetic energy and Joule losses induced by the eddy currents, as well as the equivalent complex permeability at the strand level, for configurations with the single rectangular and the hexagonal lattice respectively. At the bundle level, the FEM is further applied to compute the magnetic energy of the representative unit cell of a bundle. An iterative scheme, based on the complex permeability of the strand-level, is then used to determine the effective permeability of the bundle. The resulting two-scale simplified model enables accurate evaluation of the equivalent magnetic energy and Joule losses in multi-turn Litz wire structures. Comparative analyses demonstrate that the results are in excellent agreement with those obtained from a full-scale, fine-resolution model.

Space-time modulation systems have garnered significant attention due to their resemblance to moving-matter systems and promising applications. Unlike conventional moving-matter systems, modulation systems do not involve net motion of matter, and are therefore easier to implement and not restricted to subluminal velocities. However, canonical wave-medium interaction aspects, such as scattering and energy-momentum relations, have remained largely unexplored. In this paper, we address the aforementioned issues for three dynamic systems: moving-matter blocs, moving-perturbation interfaces and moving-perturbation periodic structures, and provide corresponding general formulations along with comparisons. Our investigation reveals the significant roles played by the "catch-up" effect between waves and interfaces. Even more interestingly, it reveals different energy and momentum exchanges between moving media and homogenized moving-perturbation structures as a result of conventional and reverse Fresnel-Fizeau drag effects.

A primary challenge in wireless power transfer (WPT) systems is to achieve efficient and stable power transmission without complex control strategies when load conditions change dynamically. Addressing this issue, we propose a third-order pseudo-Hermitian WPT system whose output characteristics exhibit a stable frequency and constant power. The frequency selection mechanism and energy efficiency of the nonlinear WPT system based on pseudo-Hermitian under the coupling mode theory approximation are analyzed. Theoretical analysis indicates that under certain coupling coefficients and load conditions, the proposed system can achieve frequency adaptation in a stable frequency mode without the need to change the circuit frequency. When the load changes dynamically, the stability of the power output is maintained using a proportional integral (PI) control strategy that only collects the voltage and current at the transmitting end, eliminating the need for wireless communication circuits with feedback from the receiving side. Experimental results demonstrate that the proposed design scheme can achieve constant power transmission when load conditions change, maintaining stable and relatively high transmission efficiency. The proposed scheme exhibits benefits in practical applications since no communication is required.

Optics related to non-Euclidean geometry has been attracting growing interest for emerged novel phenomena and the analog for general relativity, while most studies are limited to the free space on rotationally-symmetric surfaces. In this paper, we focus on the light control and ray tracing on complex surfaces filled with inhomogeneous transformation medium. Within the conformal transformation optics, focal control devices and absolute optical instruments have been extended to curved surfaces. According to the equivalence between geometry and material, the metric tensor of the curved surface and the refractive index tensor are unified as the optical metric for the Hamilton's equations of light propagation on a curved surface. By solving for ray trajectories in the local coordinate system of mesh element and illuminating the refraction between non-planar elements with discontinuous media, a mesh-based ray-tracing algorithm on curved surface with medium has been proposed to validate the light control. Our research establishes a theoretical framework for light ray control in non-Euclidean space and offers an efficient tool for ray tracing in inhomogeneous medium on curved surface.

This paper presents a modified analytical method for evaluating the apparent power in homogenized models of conductor arrays, with particular focus on structures exhibiting periodic and regular hexagonal symmetry. For analyzing the proximity effect, the extended Ollendorff formula is implemented. For evaluating the skin effect, by solving the diffusion equation with Bessel functions, the apparent power inside the conductor is obtained using the Poynting theorem. The total power of the unit cell is subsequently completed through the application of Ampere’s Circuital Law. The accuracy of the modified analytical solution is validated against detailed finite element simulations across a wide frequency range. Results show that the modified method significantly improves the prediction of reactive power, especially in the high-frequency regime, while maintaining excellent accuracy in active power estimation. The effectiveness of the method is further demonstrated in large-scale, homogenized domains composed of multiple periodic and hexagonal cells. Detailed field distribution comparisons confirm the validity of the homogenization process.