Iet Electric Power Applications最新文献

Doubly fed induction generators (DFIGs) are widely used in wind energy conversion systems due to their ability to provide variable-speed operation, offering significant advantages in energy capture. However, the presence of interturn short-circuit (ITSC) faults in the rotor windings of DFIGs poses a serious threat to their reliability and performance. Detecting such faults at an early stage is crucial for preventing damage and minimising maintenance costs. Traditional rotor interturn short-circuit fault detection methods in DFIGs often rely on extracting features from measurement or control signals using techniques such as the fast Fourier transform (FFT) to analyse their frequency components. However, these methods face challenges, especially when the generator operates near synchronous speed, as they may fail to capture subtle changes in the fault indices that are indicative of ITSCs. To address these challenges, this paper proposes a novel approach for ITSC fault detection in DFIGs operating at synchronous speed using a combination of convolutional neural networks (CNNs) and transformer architectures, especially for rotor interturn short-circuits (RITSC) due to their critical impact on system reliability. By combining these two architectures, the proposed diagnostic method significantly improves the fault detection accuracy compared to traditional approaches. The model was tested on rotor and stator current data, achieving classification accuracies of 99.01% and 95.52%, respectively. Additionally, the model demonstrated excellent robustness by achieving near-perfect accuracy (100%) under super-synchronous conditions and 98.93% accuracy at sub-synchronous speeds across varying load conditions. This hybrid CNN–transformer approach provides a robust solution for real-time fault detection in DFIGs, offering enhanced performance and reliability in wind turbine systems.

With the continuous development of new energy vehicle technology, coreless axial flux motors have garnered increasing attention due to their advantages in overload capacity and power density. However, as overload capacity and power density improve, the evaluation and enhancement of the reliability of epoxy resin-encapsulated stators have become critically important. This paper evaluates the risk of epoxy resin fracture during full-load motor operation under high-temperature conditions. Additionally, the risk of coil breakage under axial forces is assessed, considering potential epoxy resin failures. By replacing the rotor auxiliary permanent magnet materials and optimising the coil fixation areas, the reliability of the coils is improved. A prototype with adjusted parameters is fabricated and tested under various operating conditions, demonstrating the structural robustness of the motor. This study contributes positively to preventing coil fractures in coreless axial flux motors.

Low-voltage helical windings, typically constructed with multiple parallel continuous transposed cables (CTCs), often include transposition structures. These transpositions alter the axial short-circuit electromagnetic force distribution within the winding, consequently influencing its axial vibration behaviour. Although previous studies have modelled winding axial vibration using ‘mass-spring-damper’ systems and examined the effect of spacers' nonlinear characteristics, the impact of transposition-induced changes in electromagnetic force distribution on axial vibration remains unexplored. This paper addressed this gap by analysing the axial short-circuit electromagnetic force distribution in two 110 kV transformer low-voltage helical windings considering the existence of transposition structures. Subsequently, axial vibration calculations were conducted using a ‘mass-spring-damper’ model. Comparison with calculation results that neglecting the transposition structure reveals that the transposition structure alters the low-voltage helical winding’s vibration mode. Specifically, it reduces both the vibration displacement and the axial stress experienced by individual disks. The results indicate that it is necessary to consider the transposition structure when analysing the axial stability of helical windings under external short-circuit faults.

This article proposes a novel model predictive control (MPC) strategy for a triaxial servo system with the capability of suppressing multiple disturbances. The discrete state-space model is built for a triaxial servo system, considering nonlinear disturbances as constraints, and Q-learning algorithm is used to identify the parameters of the nonlinear friction disturbance model. In the proposed method, a new three-part cost function was designed. It combines tracking error and control input like the traditional cost function, and adds a disturbance term. This enables the consideration of disturbances when selecting control inputs. Additionally, the optimal control law is derived through differentiation, and its stability is proven. The experimental results show that, compared to the traditional MPC, our proposed approach offers a faster dynamic response and superior steady-state accuracy. In three-degree-of-freedom experiments, the speed ripple has been reduced by 61%, although computing time has increased by 12%.

Magnetic flux leakage and excessive heating at the generator end are significant challenges in large steam turbine design. Copper shielding is an effective strategy to reduce end-region magnetic losses and enhance heat dissipation. However, under leading power factor operation, leakage flux intensifies, increasing thermal stress. This study investigates the impact of nonuniform copper shielding thickness on electromagnetic and thermal performance using a 320 MVA steam turbine generator. A three-dimensional electromagnetic model, integrated with a fluid–solid coupling thermal model, was developed to evaluate leakage flux, eddy current losses, and temperature rise under various shielding configurations. The results show that reducing the shielding thickness in the third section increases copper shielding loss by 5.12 kW compared to the baseline. The maximum temperature of the stator winding under the Type III configuration reached 63.6°C, 0.9°C higher than the baseline, whereas the peak temperatures of the copper shields increased by 3.3°C, 5.5°C and 6.1°C, respectively. The maximum magnetic flux density was 0.92 T near the pressure finger. Experimental data showed good agreement with simulations, with discrepancies below 5%. This study provides a foundation for design refinement of copper shielding, considering performance, cost and manufacturability. Future work will introduce optimisation methods to improve shielding design.

The conventional model predictive current control (MPCC) of a three-level neutral-point-clamped (3L-NPC) inverter-fed permanent magnet synchronous motor (PMSM) drive suffers from enormous computational stress due to the evaluation of comprehensive voltage vectors (VVs) in optimisation and sensitivity to manually tuned weighting factors in multi-objective cost function. This paper presents a computationally efficient MPCC algorithm that omits weighting components. The deadbeat principle directly establishes the required voltage vector (RVV), which avoids the stator current predictions, whereas a reduced control set of six VVs is derived using the angle of the RVV. The proposed cascaded technique attains current regulation and neutral-point voltage (NPV) balancing: initially, current control is achieved via six selected VVs assessed by two cost functions, thereby reducing computational load and eliminating the weighting factor. The NPV balancing control will be conditionally activated when the NPV threshold is surpassed or a small VV is selected. This control evaluates only two opposite states of selected small VV through a specified cost function. A simple optimal duty cycle technique improves current quality, whereas a unique switching frequency reduction strategy employs continuous zero VV within adjacent sampling intervals. Simulation and hardware-in-the-loop testing have confirmed the superiority of this technique compared to conventional MPCC, demonstrating reduced complexity and a substantial decrease in total harmonic distortion (THD), torque/NPV ripples, and computational burden.

High-frequency transformers (HFTs) are critical in modern power electronics, especially for application scenarios requiring compact size, high efficiency and superior thermal stability. However, their performance is often constrained by manufacturing tolerance and material property variations. This paper presents a robust optimisation design (Rob. D) method for nanocrystalline HFTs (nanoHFTs) that considers the material uncertainty, focusing on optimising power density, leakage inductance and thermal stability simultaneously. The support vector regression (SVR) is applied to the surrogate model to replace the computationally expensive finite element analysis (FEA) during the extensive uncertainty evaluations required in MCA, whereas the nondominated sorting genetic algorithm III (NSGA-III) leverages the surrogate model to efficiently navigate the trade-offs between competing objectives under variability. The Pareto-optimal solutions achieve a 1.26°C decrease in hotspot temperature of nanoHFT. The validation via simulations and a 20 kVA prototype confirms that Rob. D reduces the standard deviations of the core uncertainty to 2.7 × 10⁻³°C, substantiating the framework's efficacy in balancing competing objectives under uncertainty. Compared to conventional deterministic optimisation design (Det. D) approaches, the proposed method demonstrates enhanced robustness by treating core dimensions and magnetic properties as statistically distributed variables. This enables optimisation of both mean performance and standard deviation of objectives, ensuring resilience against the actual manufacturing dispersions.

Recently, wireless power transfer (WPT) effectively meets the demands for distance, transfer power level, system efficiency and safety, making it highly promising for various applications. In practical applications, system performance is sensitive to the coil coupling, making reliability against coupling fluctuations a real challenge. This article focuses on the coil modelling for coils for wireless power transfer systems, of which the self-inductance, mutual inductance, B field are all taken into consideration. Besides the accurate modelling, the coil optimization is conducted for better anti-misalignment to achieve a robust stable performance. Finally, an experimental prototype is implemented, and the results validate the accuracy of the proposed model.

The synchronously rotating reference frame (SRRF)-based proportional-integral (PI) control technique has been though studied in many different inverter applications, a well-designed and clearly presented application of this technique for the three-phase, three-level, three-leg and four-wire (3P3L3L4W) grid-connected (GC) neutral point clamped (NPC) inverter has not been found. Therefore, in this study, firstly, the 3P3L3L4W GC NPC inverter is controlled with the SRRF-based PI controller. Then, to achieve an optimal artificial neural network (ANN) controller in terms of computational burden and control performance, two different ANN controllers, named ANN-1 and ANN-3, are designed with data obtained from the PI controller. The control objectives of the NPC inverter are carried out by a single ANN in ANN-1 and by three independent ANNs in ANN-3. The training results for ANN-1 and ANN-3 are approximately the same, but their computational burdens are quite different. Because ANN-3 consists of three ANNs with minimum complexity, it has much less computational burden than ANN-1. Their control performances are compared by using the MATLAB/Simulink, and presented for constant current reference, sudden changes in current reference, current reference with white Gaussian noise, sudden changes in DC source voltage, grid voltage imbalance, sag and swell, and different line filter parameters.

To address the critical challenges in defect detection of basin insulators utilised in power distribution systems, this study proposes a Dilated Weighted-Across Stages Pyramid Network-Transformer (DSPN-Transformer) framework specifically designed for x-ray digital radiography (X-DR) applications. The current methods face issues such as limited robustness to imaging artefacts, high miss rates for subtle defects, and indistinct feature representations in low-contrast regions. The proposed framework leverages a Swin Transformer backbone to model long-range dependencies while enhancing attention to subtle defect boundaries in low-contrast regions. Building upon this foundation, a new Dilated Weighted-Across Stages Pyramid Network (DSPN) is designed to dynamically adjust multi-scale receptive fields and spatial-channel weights, effectively amplifying defect-related features. Additionally, a Dynamically-Aggregated Feature Module (DAFM) is introduced to achieve adaptive channel-wise fusion of hierarchical features, further improving the discrimination of defect patterns. Experiments on the MLDB_IRD dataset demonstrate that the proposed DSPN-Transformer achieves 97.58% accuracy, 97.23% AUC, and 95.57% F1-score. The DSPN-Transformer ensures reliable operation of power systems through intelligent diagnosis of critical grid components.