Jianzhong Zhang

Accurately locating switching device faults in multilevel converters remains a challenge, particularly considering the scarcity of labeled fault data in practical industrial applications. To address this, this paper proposes a data-driven fault detection framework based on a simulation transfer learning network (STLNet). First, raw three-phase current signals are preprocessed using resampling, wavelet denoising, and normalization to generate 2D current feature images. To enrich the fault samples, a symmetry-based augmentation strategy is applied. Subsequently, a lightweight convolutional neural network is pre-trained on abundant simulation data to learn fundamental fault signatures. Finally, the designed model is transferred to the real domain by fine-tuning with a minimal amount of experimental data. Experimental validation on a T-type three-level converter platform demonstrates that the proposed STLNet achieves superior diagnostic accuracy and generalization performance compared with traditional methods, while significantly reducing the dependency on real-world fault data.

Reliable detection of the rotor winding faults in the doubly fed induction generator (DFIG) is crucial for the resilience of the variable speed energy systems. High-resistance connection (HRC) and inter-turn short circuit (ITSC) faults cause current distortions that are remarkably similar, and the rapid rotor side dynamics and the DFIG multimode operation ability also make fault diagnosis more difficult. This paper proposes a three-layer diagnostic framework named ZSC-CASI-CADI which leverages three-phase rotor currents in conjunction with rotor zero-sequence current (ZSC) for comprehensive rotor winding fault diagnosis. Fault detection is realized through ZSC magnitude and the Cosine Angle Spread Indicator (CASI) enables the strong discrimination between HRC and ITSC faults using the dispersion of rotor current phasors from the ZSC reference. Fault localization is achieved using the Current Angle Difference Indicator (CADI), which determines the faulty rotor phase through the angular deviations in rotor currents from the ZSC. The methodology is verified with extensive simulation results to demonstrate the accurate, real-time fault detection, discrimination, and localization of DFIG rotor winding faults under different load and rotor speed conditions including sub-synchronous and super-synchronous modes. The results show that the proposed framework provides a light and effective solution for rotor winding fault monitoring of the DFIG systems.

This study examines the effectiveness of negative-sequence current analysis in identifying broken rotor bars for inverter-fed induction motors under different load and speed conditions. To improve diagnostic reliability, inverter-induced harmonics in the negative-sequence current were separated using a Spectrum-Tracking Hybrid Adaptive Extended Kalman Filter method, resulting in a clearer signal representation. A thorough assessment of the fault detection accuracy and sensitivity was performed to measure fault visibility and clarity. Visual comparisons of healthy and faulty negative-sequence current signals validated comparable patterns of anomalies. The findings validate that negative-sequence current analysis, enhanced by extended Kalman filter-based harmonic isolation, is a dependable and resilient technique for detecting broken rotor bars, especially in the early stages of faults and under various operating conditions.

A third quadrant characteristic (3rd-QC) model is proposed in this article, which accurately predicts the reverse-conduction behaviors of the silicon carbide (SiC) mosfet. The proposed model considers the floating substrate effect, interface state effect, and temperature sensitive effect in detail, so the I-V trajectory from the source to the drain terminal could be efficiently evaluated. A dynamic test platform is designed to extract the 3rd-QC data more comprehensive than that provided in datasheet, thus the proposed model works well in a wide operation range of the SiC mosfet. The proposed model decouples the currents in the metal–oxide–semiconductor (MOS) path and body-diode path, making the current-sharing ratio clearly observable. In order to verify the accuracy of the proposed model, a dynamic validating platform is implemented, and the complex converter is also simulated with the proposed model. The simulation results agree well with the experimental results, demonstrating the great potential of the proposed model in predicting the 3rd-QC behaviors of the SiC mosfet.

Inter-turn short-circuit faults in induction motors might lead to overheating, torque imbalances, and eventual motor failure. This paper presents a fault detection framework for accurately identifying ITSC faults under various operating conditions. The proposed method integrates negative-sequence current analysis utilizing wavelet-based filtering and symmetrical component decomposition. A fault detection index to effectively monitor motor health and detect faults is presented. Moreover, the fault location is determined by phase angles of fundamental components of negative-sequence currents. Experimental validations were carried out for an inverter-fed induction motor under variable speed and load cases. These showed that the proposed approach has high sensitivity to early-stage inter-turn short circuits. This makes the framework highly suitable for real-time condition monitoring and predictive maintenance in inverter-fed motor systems, thereby improving system reliability and minimizing unplanned downtime.

DFIG rotor windings face high stress and transients from back-to-back converters, causing inter-turn short circuit (ITSC) faults. Rapid rotor-side dynamics, combined with the unique capability of DFIG to operate in multiple modes, make the fault detection in rotor windings more challenging. This paper presents a comprehensive methodology for online ITSC fault diagnosis in DFIG rotor windings based on zero-sequence current (ZSC) component analysis under variable operating conditions. Fault features are identified and defined through the analytical evaluation of the DFIG mathematical model. Further, a simple yet effective algorithm is presented for online implementation of the proposed methodology. Finally, the simulation of the DFIG model is carried out in MATLAB/Simulink under both sub-synchronous and super-synchronous modes, covering a range of variable loads and low-frequency conditions, along with different fault severity levels of ITSC in rotor windings. Simulation results confirm the effectiveness of the proposed methodology for online ITSC fault detection at a low-severity stage and precise location identification of the faulty phase within the DFIG rotor windings under both sub-synchronous and super-synchronous modes.

As a noninvasive detection method, speed signature analysis for servo motor bearing fault detection has attracted much attention in recent years. The selection of the optimal demodulation band is critical for accurate bearing fault detection. However, under dynamic load conditions, an incorrect band is often selected, resulting in failure of feature extraction. This article proposes an optimal demodulation band search method based on the speed signal, which is suitable for dynamic load conditions. First, the fault mechanism of the speed signal due to bearing fault excitation is studied by Laplace domain modeling, and the fault characteristic reflected in the speed spectrum is revealed. Then, a simple finite impulse response filter based on the window function method is designed under the structure of arborescent filter bank. Next, an evaluation index based on the spectral harmonic-to-noise ratio of the angular domain speed signal is proposed to determine the optimal filter parameters. Finally, bandpass filtering and envelope demodulation are performed to achieve effective feature extraction. Experimental results demonstrate the superiority of the proposed method, which has wide application prospects in the field of noninvasive safety monitoring.

High torque density permanent magnet (PM) machines have become an increasingly critical requirement for drive systems. Traditional PM machines face significant challenges in achieving higher torque densities due to limitations imposed by electrical loading and heat dissipation conditions. In this paper, a novel dual-sided consequent-pole flux-modulated PM machine (DCFPM) is proposed and compared. Both stator and rotor are adopted with consequent-pole PM arrangement to enhance modulation effect and improve torque density. Based on the air-gap field modulation theory, the topology, operating principle, and flux modulation effect of the proposed DCFPM are analyzed. The electromagnetic performances, including EMF, torque characteristics, PM utilization ratio, and power factor, are evaluated by finite element analysis. The results show that the bidirectional flux modulation in the proposed DCFPM is beneficial to enhancing flux densities and torque production capability. Finally, the discussions and conclusions about different PM machines are carried out; it can be concluded that the proposed machine with 12-slot/34-pole/12-modulation pole can offer merits of high torque density, high power factor, and excellent overload capability.