Iet Electric Power Applications最新文献

To mitigate the potential loss of computational accuracy in the Reduced-Order Model (ROM) due to modal changes during transformer operation, this paper proposes a dynamic updating method for the ROM. This method enables the model to dynamically adjust and adapt to system changes. When transformer operating conditions change, new snapshot data is employed to update the original snapshot matrix, while the POD modes are updated by integrating matrix low-rank decomposition with the Singular Value Decomposition (SVD) results of the original snapshot matrix—thus avoiding the need of SVD for the new snapshot matrix. By incorporating discrete measurement data from the winding temperature, the modal coefficients are solved in real-time based on Gappy POD, facilitating the construction of the dynamic ROM. The proposed method was validated using a simulation model of 110 kV transformer windings. The results demonstrates that the maximum error in updating the POD modes is only 3.60 × 10⁻⁶, with a single update requiring approximately 0.12s. Furthermore, the dynamic ROM reduces the maximum error by 1.78 K. Without considering the snapshot matrix formation time, the average computation time for each time step is about 0.02s. This study presents a novel solution for the dynamic application of the ROM in the transformer temperature field.

Reliable fault diagnosis of power transformers is vital for ensuring the safe and continuous operation of power systems. Although deep learning methods have shown success with single-sensor data, their diagnostic performance remains limited due to the inability to capture complex, multimodal fault characteristics. To address this, we propose an attention-guided semi-supervised vibration-acoustic fusion (AG-SVAF) model, which combines vibration and acoustic signals to enhance diagnostic robustness under limited labelled data conditions. The model integrates time-frequency representations derived via short-time Fourier transform (STFT) with a multilevel attention mechanism—including channel, spatial and self-attention—to highlight fault-relevant features and model cross-modal dependencies. A novel attention-weighted consistency loss further improves the utilisation of unlabelled data during training. Validated on practical transformer datasets, AG-SVAF achieves superior performance in terms of diagnostic accuracy and stability, particularly under challenging scenarios involving class imbalance and label scarcity. This approach provides a promising and scalable solution for intelligent condition monitoring in real-world power system applications.

One important usage of electric motors is railway transportation. There are different types of motor analyses available in railway transportation such as electromagnetic, thermal and mechanical analyses. The goals of thermal analysis are to calculate and monitor motor temperature in components, avoid motor damage, insulation break and increase lifetime. Methods of thermal analysis include: lumped parameters (LP), finite element (FEA) and computational fluid dynamics (CFD). Good analysis should be suitable for the motor type and geometry, obtain good accuracy and reduce time consumption. In this paper, the authors perform the thermal analysis of a self-ventilated 200 kW induction traction motor used in urban trains. First, the thermal network is created and the proposed LP approach is provided to calculate temperatures. This model consists of two steps which makes calculation easier, takes less time and can be used for complex parts without changing the whole model. The temperatures of important parts such as end winding can be calculated by the proposed model with simple equations, without any change in motor parameters matrix, so the proposed model is more flexible and can be used for different cases. Then, FEA and CFD are used to simulate the motor. Some model parameters are optimised using equations to make the model more accurate. Finally, simulation results are compared with experimental test results in order to validate motor performance and the proposed model results.

With the consideration of the DC-link voltage fluctuation, the motor output torque contains a large number of fluctuating components, which may trigger electromechanical coupling resonance (EMCR) and deteriorate the system control performance. In order to solve this problem, the model of the electromechanical coupled system is established, and the mechanism of EMCR generation is analysed. On this basis, a method to suppress EMCR by destroying the EMCR generation conditions is proposed. The method is implemented by stator voltage harmonic compensation and speed harmonic feedback. In addition, the control performance of the system with the proposed suppression method is analysed under different conditions. Finally, extensive tests are performed to verify the effectiveness of the proposed method.

In this study, a novel six-phase axial flux switching permanent magnet machine (AFFSSPM-TS) with a twisted structure is explored to enhance fault tolerance and overall system reliability. This design integrates the advantages of flux switching permanent magnet (FSPM) machines, such as high-power density and robust structural characteristics, while significantly improving fault resilience. The machine's electromagnetic behaviour and fault-tolerant capabilities are assessed through experimental validation. To ensure stable torque output under single-phase faults, an advanced control strategy is introduced, regulating the q-axis current and zero-sequence components within the synchronous reference frame. This approach effectively mitigates torque loss caused by phase disconnection while optimising copper losses. Furthermore, a weighted multi-objective optimisation framework, utilising a Genetic Algorithm (GA), is implemented to refine the fault-tolerant current references, maximising torque output and minimising energy dissipation under fault conditions. For short-circuit faults, a compensation strategy based on fault decomposition is developed, eliminating the need for complex real-time computations by directly compensating for faulty phase currents. Experimental results on a prototype machine confirm the proposed control strategy's effectiveness in maintaining torque performance and fault-handling capability.

Due to the inherent nonlinearity and saturation in the magnetic circuits of tubular permanent magnet linear motors, the analytical method (AM), while computationally efficient, often fails to capture complex electromagnetic behaviours accurately. In contrast, the finite element analysis (FEA) offers high precision but is time consuming. The nonlinearity of magnetic materials introduces strong input–output coupling, while saturation leads to localised deviations in field distributions, both of which reduce the effectiveness and generalisability of conventional modelling approaches. To overcome these challenges, a physics-informed, data-driven modelling approach is proposed. Initially, a novel hybrid modelling framework based on physics-informed neural networks (PINNs) is introduced. In this framework, AM is incorporated into both the input-output layers and the relevant variables, thereby enabling the direct embedding of physical constraints into the loss function. Consequently, the network's training process is rigorously guided in accordance with established physical principles. To further enhance prediction accuracy and generalisation, a transfer learning framework is integrated into PINN, utilising pre-trained datasets from AM and fine-tuning the model using high-accuracy datasets derived from FEA. Additionally, to optimise the physical information-related hyperparameters that impact model accuracy, functional analysis of variance is employed to quantitatively assess their importance and determine the optimal hyperparameter values. Experimental results show that, with training sample sizes representing only 5% of the FEA data, TL-PINN achieves significant improvements over DNN, including a 74.25% reduction in (1 − R²), a 49.51% reduction in RMSE, and a 50.46% reduction in MAE. These findings demonstrate that TL-PINN delivers superior accuracy while utilising substantially fewer FEA datasets.

The selected electromagnetic field model of a hydro generator directly affects the calculation of the no-load magnetic field, which in turn affects the grid-connected power quality of the hydrogenerator and the power transmission safety. Tubular hydro generators have a horizontal structure with less internal space than the conventional vertical hydro generator, which results in a particularly complex and strong internal magnetic field distribution. This study investigated the influence of the selected electromagnetic field model on the calculation of the no-load magnetic field of a tubular hydro generator. Straight and skewed stator slots were considered for the structure of the hydro generator. Three models were considered: the transient motion electromagnetic field-circuit coupling model, the transient motion electromagnetic field model, and the static electromagnetic field model. The calculation accuracy and computational efficiency of the three models were evaluated by comparison to measured data. The results were used to make reasonable suggestions for the selection of a suitable electromagnetic field model in different scenarios. The findings are expected to support the electromagnetic field analysis and design of hydro generators.

This paper presents two novel hybrid rare-earth and ferrite permanent magnet (HPM) asymmetric V-shape and U-shape interior PM synchronous machines (IPMSMs) with high ferrite PM (FEPM) torque contribution accounting for the enhanced demagnetisation withstand capability of FEPM at both open circuit and overload conditions. The proposed topologies are designed and compared with a rare-earth PM (REPM)-based symmetrical V-shape baseline in terms of electromagnetic performances, mechanical strength, demagnetisation withstand capability, and PMs cost. All machines are optimised for the same torque with the minimum volume of high-cost REPM at the same specification and size as a commercialised electric vehicle (EV) IPMSM. It is shown that the synergies of magnetic field shifting effect and HPM utilisation have improved the torque per REPM usage in both proposed machines. However, the magnetic field shifting of the proposed HPM asymmetric U-shape IPMSM is twice of that in the V-shape IPMSM counterpart along with a slightly better FEPM demagnetisation withstand capability. Meanwhile, the results show that the proposed HPM asymmetric V-shape IPMSM would be cheaper than the U-shape counterpart as the former and latter topologies require ∼31% and ∼23.5% less REPM volume than the baseline, respectively. Finally, two small laboratory size prototypes are made and tested to verify the finite element analyses.

With the growing share of wind power generation in the global energy structure, sub-synchronous oscillations (SSOs) triggered by the integration of wind power into the grid, as a potential dynamic instability phenomenon, have a non-negligible impact on the performance and safety of the power system and even lead to serious operational risks. Starting from the operation mechanism of the doubly-fed induction generator (DFIG) and in combination with the control of the rotor-side converter (RSC), an equivalent impedance model of the DFIG is constructed. Employing the Morris method and the extended Fourier amplitude sensitivity test (EFAST) method, a comprehensive global sensitivity analysis of the system impedance is conducted, progressing from qualitative to quantitative analysis. High-sensitivity parameters are identified, and the system dynamic interval is partitioned through the joint adjustment among these parameters. Subsequently, a cooperative optimisation method for high-sensitivity parameters is proposed to optimise the parameter set within the instability region to effectively suppress SSO. Finally, the simulation model of the DFIG grid-connected system with series compensation is established, and the feasibility of the optimisation method is verified using the Middlebrook criterion. The results demonstrate that the optimisation method exhibits strong adaptability under different operating conditions, effectively mitigating the risk of SSOs and ensuring stable operation of the wind power system.

In recent years, a significant amount of research on modelling of electrical machines was dedicated to high fidelity hybrid models that incorporate pre-calculated FEA data in the form of lookup tables (LUTs). Despite the increasing interest in this dynamic modelling approach, the literature largely disregards how the construction process of LUTs can impact both the accuracy of the model and the computational efficiency. This paper explores the LUT accuracy level attainable by application of various relevant data fitting interpolation algorithms and data calibration parameters using a standard permanent magnet machine geometry. It is demonstrated that an optimal trade-off between high accuracy LUT demand and the inherent high computational requirements associated with creating the requisite FEA datasets is important for facilitating effective development of hybrid model LUTs.