ABSTRACT This article addresses the significant limitations of conventional pulse injection for sensorless control of switched reluctance motors (SRMs) at low speed. Near the aligned rotor‐stator position, a small pulse response current leads to a low signal‐to‐noise ratio (SNR), whereas near the unaligned position, an excessive response current generates additional copper loss and negative torque. To overcome these limitations, this article proposes a dual‐threshold pulse injection method that adaptively adjusts the injection frequency and duty cycle in real‐time based on the current response constraining the peak current within a given threshold range. Furthermore, the traditional inductance vector method is highly susceptible to magnetic saturation, which restricts its application to no‐load and light‐load conditions. To address this, an incremental inductance reconstruction method based on saturation region identification is presented. This method synthesises the full‐cycle reconstructed incremental inductance by combining the extracted unsaturated incremental inductance with the reconstructed inductance of the saturated region, which is calculated based on the signal pulse width determined by the saturation region and the incremental inductance values measured at the rising and falling edges of the pulse. This reconstructed inductance exhibits high immunity to saturation effects; its substitution into the inductance vector method enables reliable sensorless operation under variable‐load and variable‐speed conditions. The effectiveness and accuracy of the proposed control strategy are validated through comprehensive simulation and experimental results.
ABSTRACT With the large‐scale integration of renewable energy and widespread deployment of power electronic devices, high‐frequency harmonics have become increasingly prevalent in power systems, which renders the impact of distributed capacitance on power transformers a critical concern for power system operation. Conventional high‐frequency models of transformer fail to adequately consider the combined effects of core saturation and distributed capacitance. This study proposes a capacitance conversion method based on the duality principle, thus establishing an enhanced unified magnetic circuit model for power transformers. The methodology references the nonlinear excitation inductance fitting approach derived from air‐core inductance to efficiently compute reluctance. The study proposes a distributed capacitance fitting function derived from transformer design procedure, which enables rapid and accurate capacitance computation. The enhanced and conventional models are developed using C++ programming and compared with the measurement results of transformer entities. Results indicate that the enhanced model more accurately characterises capacitive–inductive coupling interactions. This model provides a reliable theoretical tool for the high‐frequency transient analysis of power transformers.
The double stator wound-excited magnetic field modulation (WMFM) (DS-WMFM) machine has two stators with field and armature windings, respectively, and the sandwiched rotor with iron pieces. This paper proposes a lumped parameter thermal model (LPTM) for the DS-WMFM machine. A detailed method for calculating and constructing the LPTM for the DS-WMFM machine is given in this paper. To validate the prediction accuracy of the established LPTM, four experimental tests including DC test, rated load test, variable condition test and overload test are conducted. The experiments results show that the maximum absolute error of the temperature can be limited within 6.25 degrees C, 3.64 degrees C and 6.14 degrees C in rated load test, variable condition test and overload test, respectively.
This paper presents the design and a comprehensive electromagnetic, thermal and mechanical design assessment of a high-speed radial flux permanent magnet synchronous coupler (RFPMSC) integrated with an electromagnetic disconnection mechanism for reliable torque transmission in safety-critical applications. The main novelty of this work lies in the development of a unified electromagnetic-thermal-mechanical design framework in which torque capability, thermal behaviour, temperature-induced demagnetisation of permanent magnets, mechanical integrity, vibration characteristics and controlled electromagnetic disengagement are considered simultaneously. An analytical electromagnetic model based on a Halbach magnet arrangement is developed to estimate the transmitted torque and magnetic field distribution, and the results are validated using two- and three-dimensional finite element simulations. A multi-objective design study is conducted to investigate the trade-offs among torque capability, rotor mass, torque-to-mass ratio and manufacturing cost, leading to an optimised coupler configuration suitable for high-speed operation. The axial magnetic force acting between the rotors is also analysed, and a linear solenoid actuator (LSA) is designed to enable rapid and reliable disengagement of the magnetic coupling. Unlike most existing studies, the thermal behaviour of the coupler is explicitly modelled and quantitatively linked to permanent magnet demagnetisation and the associated reduction in transmitted torque under worst case operating conditions. The results show that the proposed RFPMSC can transmit a maximum torque of approximately 77.26 N & centerdot;m under adverse thermal conditions. Furthermore, the designed LSA can generate up to 520 N axial force, enabling reliable electromagnetic disengagement within approximately 55 ms. These results demonstrate the effectiveness of the proposed integrated design methodology for high-speed magnetic coupler systems.
Open-circuit (OC) faults in rotating diode rectifiers (RDRs) pose a severe threat to the operational safety and reliability of wound-field synchronous starter/generator (WFSSG) systems. Although existing position-sensorless diagnostic frameworks based on the rotor current vector modulus (CVM) are theoretically applicable to both rotating and stationary conditions, prior research has primarily established characteristic mechanisms, analytical relationships and threshold criteria specifically for rotating operation. Under standstill conditions, the DC components of the three-phase rotor currents do not contribute to stator electrical signals due to the absence of time-varying magnetic fields. Consequently, the DC components of the rotor current no longer contribute to the formation of stator-side signatures, rendering the CVM mathematical expressions and fault criteria established under rotating conditions no longer applicable. To address theoretical gaps under standstill conditions, this paper analyses fundamental differences in feature formation between rotating and stationary states using electromagnetic induction laws. An analytical standstill CVM model is derived within the CVM framework, and characteristic indicators with identification thresholds are established for static detection. Experimental results validate the derived standstill CVM analytical model and the proposed diagnostic indices, demonstrating that the method can accurately and robustly identify dual-diode open-circuit faults in the rotating diode rectifier under standstill conditions.
The linear primary permanent-magnet vernier machine (LPPMVM) is a promising candidate for long-stroke applications owing to its high thrust density and cost-effectiveness. Nevertheless, significant thrust ripple persists due to rich spatial harmonics in the air-gap magnetic field. To address this limitation, this paper proposes a novel linear primary permanent magnet vernier machine incorporating a reverse Halbach array (RH-LPPMVM). The core innovation lies in the synergistic use of reversed magnetisation segments and a deliberate permeance phase shift to selectively attenuate specific harmonic orders in the armature flux density distribution. Analytical and finite-element studies reveal that the 2nd and 22nd harmonic components dominate electromagnetic force fluctuations. By suppressing these critical harmonics through the proposed reverse-Halbach design, the RH-LPPMVM achieves a significant reduction in thrust ripple while maintaining the average thrust. Experimental validation on a prototype confirms the overall effectiveness of the proposed method, though discrepancies between predicted and measured ripple are discussed and attributed to practical manufacturing factors.
ABSTRACT Magnetic screw motor (MSM) is a typical double degree‐of‐freedom actuator. Inevitably, such actuators require two position encoders, including a linear encoder and a rotary encoder. To achieve the reduction of system costs and the improvement of system reliability, a sensorless control strategy based on high‐frequency square‐wave signal injection is proposed to replace the rotary position encoder. By injecting high‐frequency pulsating square‐wave signals into the d ‐axis voltage, the estimated rotor position information can be decoupled from the αβ ‐axes voltage through a signal processing module. Simultaneously, to simplify the system structure and improve the control performance, a sliding‐mode controller with an improved reaching law is utilised. The proposed sliding‐mode controller is implemented based on the mathematical model of the MSM. The given q ‐axis current can be obtained from the given linear displacement through the mathematical model of the MSM; thus, the rotor speed closed‐loop can be eliminated, and the traditional three‐closed‐loop structure can be simplified to a double‐closed‐loop structure. Moreover, the used reaching law can reduce chattering while ensuring the convergence speed. Consequently, the reliability and validity of the proposed method are verified through experimental results.
This paper proposes a multi-stage magnetic gearbox (MSMG) designed for automotive applications. Although traditional magnetic gears offer contactless operation, their single-ratio nature restricts performance across diverse driving cycles. The proposed architecture overcomes this by providing six distinct gear ratios within an integrated system. To manage frequency-dependent parasitic losses at high speeds, an adaptive axial magnet segmentation strategy is introduced, scaling from 3 to 10 segments across the stages. The system's electromagnetic and mechanical characteristics are evaluated using 3D finite element analysis (FEA), investigating torque performance, power loss distributions and axial shifting forces. Results demonstrate a peak efficiency of 96.97% in Stage 2, with efficiency remaining above 92.57% even at the highest speed stage (Stage 6). The study confirms that the proposed MSMG provides a high-torque-density (229.50Nm/L in Stage 1), high-efficiency and maintenance-free alternative to conventional multi-speed mechanical transmissions. By effectively suppressing eddy current losses through adaptive segmentation, this design ensures robust performance for high-power propulsion systems.
Deep learning-based image-to-image translation models (I2I) have shown strong potential in accelerating magnetic performance prediction of interior permanent magnet synchronous machines (IPMSMs). However, most existing approaches are limited to a single machine topology, restricting their generalisation capability. To address this limitation, this paper proposes a transfer learning enhanced conditional generative adversarial network (cGAN) for cross-topology magnetic flux density prediction across multiple IPMSM designs, including V-shaped, U-shaped and parallel-magnet configurations. The proposed method leverages pre-trained feature representations and adapts them to new topologies using a significantly reduced dataset, thereby lowering sampling requirements while maintaining high prediction accuracy. Various transfer learning strategies are systematically investigated to balance generalisation and efficiency. Experimental results demonstrate that the proposed method achieves accurate predictions with substantially reduced computational cost, highlighting its effectiveness for fast performance evaluation and design optimization.
In this paper, a fault diagnosis method based on a spatial trajectory diagnosis network model is proposed for T-type three-level inverters. By providing a rigorous theoretical derivation of the phase current distortion under switching constraints and neutral-point voltage shifts, the inherent mapping from time-domain waveform clipping to 3D trajectory deformation is revealed. A classification network model is employed to process these spatial point sets using symmetric functions for robust fault recognition. The proposed method is validated on a dSPACE experimental platform, showing that diagnostic accuracy remains above 92.5% even under 10% white-noise interference. Furthermore, a comparative study with existing methods demonstrates that the proposed approach achieves an optimal balance between hardware cost and diagnostic depth, identifying both single- and multi-switch faults within 16.2 ms without requiring additional voltage sensors.
Current methods for multi-characteristic jointly evaluating the ageing state of insulation paper mostly have the problems of single evaluation indicator (based solely on content) and fixed weight factors. Fixed weight factors cannot accurately reflect how the relative importance of each characteristic varies with ageing. Therefore, the applicability of the evaluation methods needs to be improved. This paper proposed a multi-characteristic joint evaluation method on the ageing state of transformer insulation paper based on multi-indicator. Methanol (MeOH), furfural (2-FAL), carbon monoxide (CO), and acidity were selected as evaluation characteristics. Content, increase rate, and evaluation error were selected as evaluation indicators. A weight model was constructed based on the CRITIC method, and the Bayesian Optimisation Algorithm (BOA) was further utilised to optimise the weight factors, thereby establishing a multi-characteristic joint evaluation model. The optimal weight values for each characteristic were determined through an accelerated thermal ageing test conducted at 393.15 K. The experimental results showed that the relative evaluation errors of the proposed method were within 6.81%, which was significantly lower than those of the single-characteristic evaluation methods and demonstrated better applicability. These results verified the effectiveness of the proposed method.
By combining the advantage of yokeless and segmented armature (YASA) machine and flux modulation effect, the flux modulation YASA (FM-YASA) machine exhibits superior torque density and efficiency. The finite element (FE) method is the most accurate way to predict the machine performances. However, it will cause a heavy computational burden. Therefore, a magnetic equivalent circuit model (MECM) is proposed in this paper to analyse the FM-YASA machine rapidly, which is 498 times faster than three-dimensional FE analysis and 3.4 times faster than two-dimensional FE analysis approximately. Compared with traditional lumped-parameter MECM, the tiny rectangular grids are adopted for the refined modelling of airgap, permanent magnet (PM), modulation slot and stator slot opening. The refined magnetic grids of airgap and PM unify the model to achieve a dynamic MECM which can auto-update connection of the magnetic reluctance during rotor rotation. Since PM with stepped slots is designed to facilitate assembly and enhance mechanical strength, the leakage flux of PMs is captured accurately by the refined PM model. The effectiveness of the proposed MECM is verified by the FE analysis and experiment, with errors < 5%.
Dissolved Gas Analysis (DGA) serves as the technological foundation for the primary diagnostic method used to detect incipient faults in transformers, with traditional approaches such as the Duval pentagon, Duval triangle, and IEC ratio methods remaining widely adopted in practice. These approaches are fundamentally limited by their deterministic classification thresholds and compromised diagnostic precision. Therefore, this study proposes a transformer defect diagnosis approach utilising traditional DGA feature similarity analysis. First, the collected transformer fault data are mapped in Duval pentagon, Duval triangle and IEC ratio method-based three-dimensional coordinates. Ordering Points to identify the clustering structure (OPTICS) is employed to identify fault clusters corresponding to distinct fault types and compute their respective centroids. Second, this study derives a novel discriminative feature based on the Euclidean distance between the acquired DGA data and each fault cluster centroid. The derived feature combination serves as input features for the eXtreme Gradient Boosting (XGBoost) model. Verification shows that the proposed transformer fault diagnosis method achieves a mean classification accuracy of 92.8%. Finally, SHapley Additive exPlanation (SHAP) is employed to evaluate the contribution of different feature indicators to the model results. The proposed model is validated through a field case study, demonstrating both feasibility and explainability.
The development of power electronic converters operating at cryogenic temperatures is limited by the lack of data on electronic components outside the standard temperature range specified by manufacturers (typically above -65 degrees C, i.e., 208 K). To fill this gap, the construction of an open access database of power electronic component characteristics at cryogenic temperatures is presented. As a first step, the focus is on the electrical characterisation of power diodes at both room (293 K) and liquid nitrogen (77 K) temperatures. We present a tailored experimental protocol that allows their I-V characteristics to be obtained in a reproducible manner at both temperatures. Additionally, we share the results obtained for more than 60 power diodes, having nominal current from a few amperes to several thousand amperes. We also analyse the variability among same-reference diodes and the impact of repeated thermal cycling on a given diode. Finally, we demonstrate the practical usefulness of such database through a case study. We predict the performance (losses, efficiency) of a 1 kW single-phase diode rectifier at 77 K using only the I-V characteristics of the diode from the database, and we show that it matches the performance experimentally measured on the corresponding prototype.
The trend towards electrifying transportation systems has stimulated research endeavours aimed at developing electric machines that are not only high speed and efficient but also low-cost and compact. To support this transition, the US Department of Energy (DOE) has established ambitious targets of $7/kW cost and 12 kW/L power density for the electric drivetrains. This paper attempts to meet these targets by proposing high-speed permanent magnet assisted synchronous reluctance machine (PMASynRM) topologies enabled by hybrid magnet strategies, combining rare-earth (RE) magnets with low-cost, RE-free alternatives. Modified 2-layer and 3-layer U-shaped rotor configurations, featuring structural reinforcements to withstand high mechanical stress at elevated speeds, are developed. A multi-objective design optimisation framework is employed to optimise the rotor design, targeting reductions in RE magnet volume, magnet demagnetisation risk and torque ripple while achieving the desired electromagnetic performance. A comprehensive analysis of the optimised rotor designs, compared against a 16,000 rpm, 800 V baseline IPM motor, shows over 70% RE reduction, demagnetisation risk below 3.5% and torque ripple under 10%. Detailed analyses of performance trends across various magnet combinations and rotor configurations highlight the viability of hybrid magnet PMASynRM designs as cost-effective, robust and energy-efficient solutions for next-generation EVs.
The voltage drops in the distribution network is one of the most important power quality phenomena. At the voltage-sensitive feeders, voltage drop even for a short period of time causes load interruption, therefore local solutions are necessary to improve voltage drops in these feeders. Dynamic Voltage Restorer (DVR) is used as a series compensation device in power distribution networks to eliminate instantaneous voltage drops. In this paper, a PI controller is improved for DVR to compensate effectively voltage drops and reduce the series transformer inrush current. The controller parameters are optimised using the modified arctic puffin optimisation (MAPO) algorithm in order to improve the performance of DVR. A medium voltage network with a voltage-sensitive load is simulated in MATLAB environment. Besides, balanced and unbalanced voltage drops are considered in the presence of DVR. The simulation results show that the proposed method compensates balanced and unbalanced voltage drops quickly and effectively. At the same time, significantly reduces the series transformer inrush current. The harmonic distortion of the load voltage is also reduced in the presence of the proposed DVR and the power quality is improved. Also, the proposed results are compared with the other algorithms and references results.
The split ratio of an axial-flux permanent-magnet (AFPM) machine, defined as the inner-to-outer stator diameter ratio, is optimised to maximise electromagnetic torque under thermal constraints considering both copper loss and iron loss. An analytical relationship between torque and split ratio is derived, and the predictions are validated by finite-element analysis (FEA). The results show that, when stator iron loss is taken into account, the achievable torque is reduced and the optimal split ratio shifts compared with the copper-loss-only case. The influences of air-gap length, flux-density ratio and operating speed on the torque-split-ratio relationship are further quantified, and the effect of slot shape is also examined. Three-dimensional FEA and experimental results confirm the predicted trends and optimal points, demonstrating that the proposed method can serve as a fast preliminary sizing tool for thermally constrained YASA-type AFPM machines.
Model predictive control (MPC) provides notable advantages over conventional linear pulse width modulation (PWM)-based switching techniques, including faster dynamic response, the ability to manage multiple inputs and outputs simultaneously and the ability to integrate various constraints. Despite these advantages, the use of MPC algorithms in specific multilevel inverter (MLI) topologies remains challenging and underexplored, particularly due to issues such as computational complexity and the need for real-time processing capabilities. Considering the advantages of MLIs in the solar PV system, it is necessary to analyse the MPC benefits for the same. Therefore, a finite control set model predictive control (FCS-MPC) algorithm for a nine-level switched capacitor-based quadruple-boost multilevel inverter (QBMLI) is proposed in this paper to exploit the benefits of MPC and MLI in the solar PV system. For this, the system model was initially developed in MATLAB/Simulink, focusing on a PV-fed QBMLI with an RLE load, which mimics the characteristics of the motor load. Furthermore, a comparative analysis of FCS-MPC with the linear PI-based level-shifted PWM (LSPWM) technique is presented to validate the superior performance of the FCS-MPC for the considered nine-level QBMLI. The performance of the system is then validated in real time using a dSPACE Scalexio-based real-time hardware-in-the-loop (HIL) setup. Key performance metrics, including load current tracking, transient response and harmonic distortion in load current and voltage, are evaluated under different sampling intervals. In addition, a detailed analysis is carried out to assess the influence of load changes on system performance, providing essential information for the practical application and optimisation of FCS-MPC.
The limited accuracy and long computation time of temperature-field solutions for oil-immersed transformers impede rapid perception of their internal state. This paper proposes a tailored hybrid physics-data surrogate model for steady-state temperature fields of oil-immersed transformers that deeply integrates Proper Orthogonal Decomposition (POD) with a specifically modified neural network architecture. First, a full-order model of the oil-immersed transformer is built, and temperature-field simulations are performed under various operating conditions. Second, Proper Orthogonal Decomposition (POD) is applied to extract characteristic modes from the simulated temperature-field snapshots, transforming the nonlinear temperature field into a linear combination of modes and modal coefficients. Finally, a Gaussian-activated fully connected neural network with local approximation capability is constructed, and a physics-informed loss function embedding POD modal importance is designed to approximate the nonlinear mapping between transformer operating parameters and modal coefficients, thereby completing the surrogate model construction. For the temperature-field solution under a given condition, the field can be rapidly reconstructed simply by linearly weighting the modes with modal coefficients predicted by the surrogate model. Validation shows that the proposed surrogate model can compute a single steady-state temperature-field distribution within 0.0129 s, and the maximum computation errors for global temperature and hotspot temperature do not exceed 3.5%, with the numerical results further validated by optical fibre temperature measurement experiments, meeting the requirement for real-time temperature-field perception in on-site operation and maintenance.
This paper compares several self-cooled permanent magnet machines having different winding configurations, including concentrated single-layer and double-layer windings as well as distributed winding. The self-cooling capability is achieved by integrating fan blades into the rotor hub, specifically, the space between the rotor core and the shaft. As rotor rotates, the fan blades draw air in through the inlets and expel it through the outlets, thereby facilitating the removal of heat generated in the windings and rotor-mounted permanent magnets. The study focuses on evaluating the maximum temperatures at key machine components, such as the magnets and end windings, under varying rotational speeds. Two scenarios have been considered in this study. One assumes that all the machines have the same copper loss density while with the core losses and magnet eddy current loss being neglected, and the other uses real losses obtained by finite element models. By conducting thermal analyses under these conditions, the self-cooling capability of permanent magnet machines with different winding structures can be comprehensively assessed. Thermal analyses based on computational fluid dynamics modelling are performed under these conditions to comprehensively assess the self-cooling capability of permanent magnet machines with different winding structures. The results demonstrate that concentrated windings gain greater benefit from the proposed self-cooling scheme than distributed windings, with the double-layer configuration showing the best overall performance. The computational fluid dynamics models were subsequently validated through a series of experiments.
