In China, the large-scale renewable energy bases are typically constructed in northwestern deserts and Gobi regions with altitudes exceeding 4000 m, where high flux of cosmic rays seriously threatens the safe operation of insulated gate bipolar transistors (IGBTs) in high-voltage direct current (HVDC) valves, particularly due to irradiation-induced single-event effects (SEEs). Previously, we experimentally investigated radiation resistance of trench-gate IGBTs and analyzed their irradiation responses using commonly adopted half- and single-cell TCAD simulations. However, these simulations often overestimate the electrothermal response, leading to significant discrepancies with experimental results. In this article, a theoretical analysis of the irradiation responses of trench-gate IGBTs is conducted via multicell simulations, the results of which are exactly consistent with experimental failure behaviors. It is revealed that the irradiation-induced reconstruction of electric field at the bottom of the trench-gate triggers single-event gate rupture (SEGR), which is the primary contributor to the irradiation-induced failure in trench-gate IGBTs. In addition, current commutation and local high-temperature point transfer between adjacent cells are observed during irradiation. Consequently, the highly localized transient current and high electric field induce a rapid temperature rise that may degrade the p-n junction, but is insufficient to cause single-event burnout (SEB) in trench-gate IGBTs under typical operating condition (i.e., LET = 10 MeV center dot cm(2)/mg and V-CE = 2800 V). On this basis, the influence of dummy-gate design on radiation resistance is also analyzed to facilitate the improvement of overall device performance in future designs.

Magnetic pole shifting (MPS) can effectively reduce torque ripple in permanent magnet synchronous motors. However, its effects on electromagnetic (EM) vibration have not been thoroughly studied. This study shows that the odd-order radial EM force introduced by MPS can aggravate vibration. To address this issue, a correlation model of harmonic parity between magnetomotive force (MMF) harmonics and radial EM force harmonics is established to reveal the radial EM force generation mechanism in MPS motors. Based on this model, an axially staggered pole shifting structure is proposed. Adjacent rotor segments generate even-order flux density harmonics with opposite phases, producing axially opposite odd-order radial EM forces that cancel each other and reduce vibration. Multiphysics finite-element analysis compares the EM and vibrational performance of the conventional and MPS motors. The proposed structure achieves improved suppression of both torque ripple and EM vibration.

Conventional fixed-gain PI phase-locked loops (PLLs) used in sensorless drive control of permanent magnet synchronous motors (PMSMs) often suffer from slow dynamic response and poor parameter robustness. An improved finite position set PLL (FPS-PLL) which may achieve high dynamic response with low computational burden is proposed in this paper. By calculating the initial iteration value through the Taylor expansion of the arctangent function, the range of initial iteration errors is narrowed, thereby reducing the number of iterations required. The rotor position is then efficiently estimated by Newton iteration. Position estimation accuracy achieves 4.96e-8 rad after just only one iteration. The proposed method provides superior dynamic performance and low computational load without parameter tuning of PLL. The performance of the proposed FPS-PLL have been experimentally investigated and compared with that of the Binary-FPS-PLL and classical PI-PLL using a 1.5 kW interior PMSM (IPMSM). The results of experiment demonstrate the effectiveness of the proposed FPS-PLL under various operating conditions.

The reliability of high-speed permanent magnet machine (HSPMM) applied to transportation electric propulsion is limited by the temperature rise of permanent magnet (PM) caused by the rotor eddy current loss (RECL). An effective rotor cooling scheme depends on the accurate calculation value of RECL. However, it is time-consuming to analyze the rotor eddy current with three-dimensional (3-D) distribution characteristics. In this paper, to balance the efficiency and accuracy of calculating the RECL, a 2.5-D model for RECL calculation based on finite element method is proposed. Firstly, taking a 24-slot, 2-pole HSPMM as an example, the characteristics of the asynchronous flux density acting on the rotor are analyzed, considering the effects of stator slotting, winding connection and current harmonics. On this basis, the 2.5-D model composed of the 3-D end model and 2-D middle model is established to consider the end effect of field and end circumferential component of eddy current, and thereby the RECL calculation procedure is formed. Moreover, the RECL of the example HSPMM is calculated using the proposed method. Finally, a small-scale locked-rotor rig for RECL test is employed and a prototype performance test is carried out, verifying the rationality and accuracy of the proposed method.

Conventional wind turbine drive trains suffer from gearbox reliability concerns and full-scale converter losses, while direct-drive solutions lead to bulky generators. This work investigates a front-end speed-flexible drive train featuring a dual-stage magnetic gear (MG) and a synchronous generator. A fixed first-stage MG and a variable second-stage MG enable electromagnetic speed regulation before the generator. By jointly selecting the two MG ratios and the generator pole pair number, the generator can operate at a constant synchronous speed and directly supply $50-\text{Hz}$ grid power without a full-scale converter. A system-level model is established that incorporates the discrete nature of MG ratios, grid-synchronization constraints, and torque and thermal limits. A discrete multiobjective design framework is formulated to minimize magnet volume and diameter while preserving efficiency. A 5-MW case study ($\mathbf{1 8 ~ r p m}$ rotor) shows that $G_{1}=12, G_{2, \text { nom }}=2.3148$, and a $\mathbf{6}$-pole-pair generator yield a $\mathbf{5 0 0 - r p m}$ synchronous speed with adequate regulation range and torque margin. Results indicate that the proposed approach provides a compact and efficient alternative for multi-megawatt drive trains, and it can serve as a practical screening tool prior to detailed finite-element and thermal verification.

As the scale of wind farms (WFs) and the number of wind turbines (WTs) continue to increase, the calculation efficiency of centralized active wake control (AWC) methods can not satisfy the real-time performance requirement. To address this issue, a dual-time-scale dynamic AWC strategy for large-scale WF is proposed in this paper. First, a centralized AWC model is developed which incorporates a dual-time-scale control strategy. The optimal yaw angle is obtained by long-term scale stochastic optimization under the time-varying wind conditions to maximize the power generation potential of WF, and different WF operating modes can be achieved by short-term scale dynamic axial induction control. Then, a spectral clustering-based WF partitioning method is introduced. The wake coupling can be obtained by engineering wake model, and the WF can be divided into multiple subregions through efficient matrix computations and eigenvalue analysis. Finally, in order to improve the calculation speed, the centralized AWC optimization problem is decomposed into several smaller subproblems that can be solved in a distributed manner based on the partitioning results. Simulation results demonstrate that the proposed method obtains reasonable WF partitioning, owns excellent power generation and fatigue load suppression performance under time-varying wind conditions, and significantly improves calculation speed.

High-reliability double-sided ring collector systems have been widely implemented in offshore wind farms (OWFs). It is challenging to achieve a globally optimal network topology and a cable capacity rating for the OWF collector system (CS) simultaneously. This paper proposes an optimal collector system planning (CSP) method for OWF with double-sided ring topology based on bidirectional flow conservation method to minimize cable costs and total power losses. By analyzing the power flow direction after faults, all fault scenarios are summarized into two fault conditions. The bidirectional flow conservation method is developed to reveal the matching mechanism between different cable sequence positions and their optimal ratings, considering the minimal rating requirements. The complex high-dimensional CSP problem, which involves the coupling characteristics of different cable parameters and system power flows, is convexified by equivalent alternative methods into a mixed-integer quadratic programming (MIQP) to guarantee a global optimal solution within feasible computation time, improving the solvability and practicality. The effectiveness of the proposed optimal CSP method has been validated in MATLAB.

When a high temperature superconducting DC-carrying coil is exposed to an AC field, AC losses occur due to vortex motion within the tapes. Therefore, 3D homogenized models are essential for the efficient evaluation of such losses in asymmetric and complex coils. However, conventional lognormal homogenization (LH) techniques typically rely on rigid division strategies, thereby failing to capture local variations in current density. Such limitations compromise the predictive accuracy of AC loss calculations in large-scale applications. This paper proposes a 3D self-adaptive homogenization (SAH) method to improve model accuracy. The model constructs a parabola-based numerical boundary to define a gradient auxiliary function, facilitating adaptive subdivision of the FEM model. Such a strategy ensures a smooth and physically consistent representation of the current density distribution across the coil cross-section. Experimental results show that the accuracy of AC loss calculation is significantly enhanced compared to the LH method. Furthermore, the adaptive iterative SAH model remains robust under high excitation amplitudes by dynamically refining the parabolic boundary solution. This capability allows the model to overcome the geometric constraints inherent in standard homogenization techniques. Overall, the proposed method provides a reliable tool for optimizing the electromagnetic performance of large-scale superconducting systems.

The randomness and intermittency of wind pose a challenge to the accurate prediction of wind power. To enhance the accuracy and efficiency of wind power prediction, this paper proposes a prediction model based on Fourier graph neural network (FGNN) and deep regret analysis generative adversarial network (DRAGAN) to accurately predict the output power of multiple wind turbines. In the proposed model, the historical power generation sequences of multiple wind turbines are embedded in the form of a hypergraph, and then the Fourier graph operators are utilized to capture complex unified spatiotemporal correlations in the Fourier space. Additionally, the historical values of each wind turbine are used as conditional inputs for the generator and discriminator, facilitating more precise generation of output features. A feature matching mechanism is introduced in the generator to enhance training stability. Through adversarial training between the generator and discriminator, the generator parameters are optimized, leading to improved prediction accuracy. Experimental results demonstrate that the proposed model achieves R-squared (R2) values of 0.9943, 0.9781, 0.9939, and 0.9858 for single-step predictions in the four seasons, showcasing outstanding multi-temporal scale predictive performance.

Variable-gear-ratio magnetic-geared machines (VGRMGMs) which could achieve adjustable gear ratios, shows substantial application prospects in driving systems. However, the VGRMGMs suffer from dynamic performance deficiencies such as slow dynamic responses and excessive speed overshoots. A robust sliding mode control (RSMC) strategy based on a load torque observer is proposed in this paper in order to improve the dynamic performance of the VGRMGM. A sliding mode speed controller and a torque observer for feedforward compensation in the current loop, is employed to improve the robustness and dynamic performance of the VGRMGM. Under speed-regulation conditions, the RSMC strategy achieves a 6% reduction in overshoot and a 25.4% improvement in response speed compared with conventional PI control. For load-variation conditions, the overshoot is reduced by 5%, and the response speed is improved by 32.1%, highlighting the VGRMGM superior transient performance. Experimental evaluations on a purpose-built test rig validate the enhancement of the RSMC strategy in terms of improved dynamic performance.

In consequent-pole dual-permanent magnet (DPM) machines, the decrease in permeability at the edges of the iron poles intensifies flux leakage, a problem that is especially pronounced in axial-flux machines. To address this, an axial-flux DPM (AFDPM) machine and a corresponding analytical method are proposed. By establishing a differential magnetic circuit model that combines both circumferential and radial directions, multiple leakage-flux effects are comprehensively incorporated. Comparative results show that the proposed machine achieves approximately a 22% increase in torque compared with dual-rectangular machines. Prototype experiments further validate the correctness of the theoretical analysis and the effective improvement in machine performance.

In China, the large-scale renewable energy bases are typically constructed in northwestern deserts and gobi regions with altitudes exceeding 4000m, where high flux of cosmic rays seriously threatens the safe operation of IGBTs in HVDC valves, particularly due to irradiation-induced single-event effects (SEE). Previously, we experimentally investigated radiation resistance of trench-gate IGBTs and analyzed their irradiation responses using commonly adopted half- and single-cell TCAD simulations. However, these simulations often overestimate the electrothermal response, leading to significant discrepancies with experimental results. In this paper, a theoretical analysis of the irradiation responses of trench-gate IGBTs is conducted via multi-cell simulations, the results of which are exactly consistent with experimental failure behaviors. It is revealed that the irradiation-induced reconstruction of electric field at the bottom of trench-gate triggers single-event gate rupture (SEGR), which is the primary contributor to the irradiation-induced failure in trench-gate IGBTs. Additionally, current commutation and local high-temperature point transfer between adjacent cells are observed during irradiation. Consequently, the highly localized transient current and high electric field induce a rapid temperature rise that may degrade PN junction, but is insufficient to cause single-event burnout (SEB) in trench-gate IGBTs under typical operating condition (i.e., LET=10MeV·cm2/mg and VCE=2800V). On this basis, the influence of dummy-gate design on radiation resistance is also analyzed to facilitate the improvement of overall device performance in future designs.

When a single open-phase fault occurs in a dual star-delta windings permanent magnet synchronous motor (DSDW-PMSM), the yaxis current in the maximum torque fault-tolerant strategy is not decoupled from the currents on the α-axis and β-axis. As a result, the conventional hybrid current fault-tolerant strategy across the full torque range cannot achieve minimum copper loss output for this type of motor. To address this issue, this paper proposes a multi-segmented hybrid current fault-tolerant strategy, which ensures that the motor can output across the full torque range while minimizing the total copper loss generated by the windings. Furthermore, finite element simulation analysis is conducted to compare the torque ripple and radial electromagnetic force density of the existing hybrid current fault-tolerant strategies. Finally, tests are conducted on a 24-slot/10-pole DSDW-PMSM with a 15° phase shift to verify the feasibility of the proposed strategy.

This article proposes a torque ripple reduction method for permanent magnet (PM) machines based on a new current harmonic injection strategy. Unlike conventional approaches focus solely on torque ripple suppression, this work proposes an optimal trade-off strategy aimed at minimizing the total system loss-encompassing both copper and core losses. The core of the proposed method lies in analytically determining the amplitudes and phases of the injected currents required for torque ripple compensation under two distinct constraints: minimum copper loss and minimum core loss. Subsequently, a weighting coefficient, adjusted according to the motor speed and operating condition, is applied to achieve high efficiency across the entire operating range. Moreover, a novel online torque ripple observation algorithm is introduced to facilitate practical implementation. The effectiveness of the proposed methodology, especially in balancing torque smoothness against loss reduction, is validated through finite element analysis (FEA).

In sensorless interior permanent magnet synchronous motor (IPMSM) drives, active disturbance rejection control (ADRC) is used to estimate the back-electromotive force (EMF) for rotor position. Nevertheless, in the existing ADRC, the aperiodic and periodic disturbances in the motor currents will significantly deteriorate the back-EMF estimation accuracy, and thus distort the estimated rotor position. This article proposes a discrete-time improved linear ADRC (ILADRC), which is composed of adaptive harmonic filtering extended state observer (AHFESO), linear ESO, and a proportional current controller. The designed AHFESO adopts least-mean square algorithm, and is capable of suppressing the parameter mismatch and harmonic disturbance simultaneously. Another linear ESO is used to obtain back-EMF for position estimation. Furthermore, the disturbance suppression capability and stability of the proposed method are discussed in discrete-time domain. Finally, the effectiveness and feasibility of the proposed method is certified on a 1.5 kW IPMSM platform. Comparative results show that the proposed ILADRC-based sensorless strategy guarantees harmonic suppression in current control and back-EMF estimation, and superior rotor position estimation ability in sensorless IPMSM drives is proved.

Floating offshore wind farms are characterized by long offshore distances and poor transportation accessibility, which result in higher failure loss cost under fault conditions in existing radial networks. Therefore, it is urgent to propose a more reliable ring topology planning scheme for the collection system. To solve the problem, a two-stage joint planning method (TJPM) is proposed for floating offshore wind farm collection system to enhance the operational reliability and reduce the full life cycle cost. In the first stage, the offshore substation location is optimized by the covariance matrix adaptation evolution strategy (CMA-ES), which performs an adaptive search by continuously and dynamically adjusting its probability distribution parameters. The optimal offshore substation location obtained in the first stage can be jointed as the constraints of the relationship optimization between wind turbines (WTs) and offshore substation areas in the second stage. In the second stage, the relationship is optimized based on a tabu search algorithm, which is customized by integrating neighborhood operations such as reallocation and swapping on key WTs. The continuous Hopfield neural network is used for cable topology planning by linearizing the full life cycle cost of floating offshore wind farms, thereby improving the computational speed in the whole topology process of the collection system. Case studies demonstrate the superiority of the proposed joint planning method for floating offshore wind farms in operational reliability and economic performance.

This paper proposes a novel multiphysics topology optimization (TO) method to enhance the material utilization rate of permanent magnet synchronous motors (PMSMs). At fisrt, the TO model considering electromagnetic torque and structural strength performance for inner-mounted PMSM rotors is established based on the solid isotropic material penalization (SIMP) method. Then, to account for the effects of magnetic saturation characteristics, a magnet equivalent circuit (MEC) model is proposed to consider the most important parts, i.e., the rotor, PM, and the airgap. Next, the electromagnetic-structural properties of the inner-mounted PMSMs are iteratively involved in the proposed unit sensitivity analysis to obtain the optimized novel rotor structure. Finally, the proposed optimal design is validated on a prototype of a 3 kW PMSM to demonstrate the effectiveness and accuracy of the proposed multiphysics TO method.

Ultralocal models (ULMs) are commonly used in model-free predictive control (MFPC) for motor drives to enhance their robustness. However, a key challenge arises from the selection of gains in ULMs, which are difficult to determine, especially for complex ULMs with multiple gains. To address this challenge, a data-gradient-based estimation is proposed in this article for optimizing ultralocal MFPC on permanent magnet synchronous motor (PMSM) drives. Given that complex ULMs incorporate multiple gains, a specific mechanism is designed to estimate these gains and lumped variables at the same time, using only the gradients of sampled and past data. To ensure stability and operating performance, a group of boundaries is inserted to limit the changing amplitudes and slopes of the estimated parameters. For demonstration, an advanced ULM (AULM) with two gains is used as an example. Experimental results confirm the effectiveness of the proposed method, as well as its advantages in terms of a simpler estimator, bounded update behavior, and implementation convenience, while achieving operating performance comparable to that of the benchmark methods.

During the missing-set operation of a dual-three-phase permanent magnet synchronous motor (PMSM), uneven circumferential distribution of winding currents leads to increased harmonic components in the air-gap magnetic field, resulting in higher losses and torque ripple. To address this issue, this paper establishes a harmonic calculation model for the winding magnetomotive force (MMF) of a dual-three-phase motor under missing-set operation based on multi-phase motor winding theory. The phase-shifting winding method is employed to exhaustively search for and optimize the MMF harmonics under different winding phase-band distributions during missing-set operation. Finally, the effectiveness of the optimization method is verified through finite element calculations. The proposed method for analyzing and suppressing winding harmonics can serve as a reference for the rapid design of multi-phase fault-tolerant motors.

Although magnetic pole shifting effectively mitigates cogging torque in Permanent Magnet Synchronous Motors, the consequent excitation of parasitic odd-order radial electromagnetic force harmonics deteriorates vibration performance. To address this issue, this article proposes a novel rotor topology that utilizes axially segmented staggered pole shifting and skewing. An analytical framework is established to determine optimal shift and skew angles based on harmonic interaction principles, specifically aiming to mitigate the dominant radial electromagnetic force harmonics. Finite Element Analysis (FEA) confirms that the proposed design simultaneously suppresses torque ripple and electromagnetic noise, exhibiting superior NVH performance compared to conventional pole-shifted machines.