Journal of Power Electronics最新文献

Aiming at the problems of high similarity and difficulty in extracting the fault features of power-switching tubes, as well as the high complexity of fault diagnosis models, the large number of parameters, and the long fault diagnosis time of the multilevel cascaded H-bridge inverter in medium-voltage and high-voltage applications, this study proposes a fault diagnosis method based on a lightweight shuffle–SimAM network. First, the proposed method establishes a lightweight parallel ShuffleNet network model and utilizes multi-sensor data as the input of each parallel network for the initial extraction of similar fault features. Second, a feature fusion module is constructed inside the network to weight and fuse the features extracted at each level of the parallel network. Then the fused features are successively advanced to further enhance the extraction of similar fault features. Finally, to maintain the network with high diagnostic accuracy while improving the level of lightweighting, deep separable convolution and SimAM parameter-free attention mechanisms are introduced into the diagnostic network. Experimental results show that the proposed method effectively reduced the complexity of the model and the diagnosis time while maintaining a high diagnosis accuracy.

This paper proposes a novel sorted level-shifted U-shaped carrier-based pulse width modulation (SLSUC PWM) strategy combined with an input power control approach for a 13-level cascaded H-bridge multi-level inverter designed for grid connection, specifically tailored for photovoltaic (PV) systems, which avoids a double-stage power conversion configuration. In this methodology, every inverter generates a quasi-square output voltage waveform with a width that is intricately linked to the output power of its corresponding PV panel. The application of this SLSUC pulse width modulation technique with input power control in a solar energy-based 13-level grid-tied inverter facilitates precise maximum power point (MPP) tracking for each of the PV panels under uniform and non-uniform irradiation conditions and ensures the consistent maintenance of capacitor voltage balance. Moreover, this novel SLSUC PWM method for 13-level inverters offers a range of benefits, including a low total harmonic distortion (THD) in the output voltage of the multi-level inverter and higher inverter and MPPT efficiencies over the existing PWM techniques. To verify the efficacy of the proposed control method over existing techniques, a PV-based grid-connected multi-level inverter with the proposed control strategy undergoes modeling and simulation using MATLAB/Simulink. Then, experimental hardware-in-the-loop (EHIL) testing is conducted to confirm and evaluate its effectiveness.

Aiming at the problem where the transmission power and transmission efficiency are affected by frequency detuning in wireless power transfer (WPT), a frequency tracking method using current-over-zero comparison and a digital control method using a dynamic time lag is proposed based on the digital signal processing (DSP) technology. The proposed method only needs to determine the time when current passes the zero point and start current sampling. Then a simple digital phase-locked loop (PLL) can be accomplished by comparing the current sampling value with the setting current value at the zero point. Combined with voltage–current double closed-loop phase-shift control, constant current or voltage output is realized. Finally, a DSP-based experimental setup with an output power of 300 W is designed for a series of verification experiments. The findings show that the method can achieve fast frequency tracking as well as phase adjustment when the air-gap distance and load are changed. It also satisfies constant current or voltage output. More importantly, the transmission efficiency of the weak inductive state achieved by the proposed method is better than that of the fixed-frequency and resonant-frequency state. In addition, it is 5% higher than that of the fixed-frequency state.

Due to the rapid advances in renewable energy technologies, the growing integration of renewable sources has led to reduced resources for Fast Frequency Response (FFR) in power systems, challenging frequency stability. Photovoltaic (PV) plants are a key component of clean energy. To enable PV plants to contribute to FFR, a hybrid energy system is the most favorable candidate, and its power sharing algorithm significantly influences the FFR capability of PV plants. In this study, a model is established for a Virtual Synchronous Generator Hybrid Energy Storage System (VSG HESS). In addition, the mechanism by which PV plants participate in fast frequency regulation is analyzed. Subsequently, a novel multi-dimensional time filtering algorithm is proposed to overcome the problems associated with the short frequency sampling periods and insufficient measurement data in PV plants. Specifically, the techniques of Multi-Delay embedding Transform (MDT), Tucker decomposition, and Multivariate Variational Modal Decomposition (MVMD) are integrated into a unified framework for improved frequency resolution prior to frequency division. Finally, the effectiveness of the proposed method is validated through online simulations performed on a VSG-controlled PV storage microgrid platform. Simulation results reveal that the proposed method is able to outperform conventional filtering algorithms in terms of frequency division accuracy and calculation speed.

Hybrid energy storage systems (HESSs) play a crucial role in enhancing the performance of electric vehicles (EVs). However, existing energy management optimization strategies (EMOS) have limitations in terms of ensuring an accurate and timely power supply from HESSs to EVs, leading to increased power loss and shortened battery lifespan. To ensure an accurate and timely power supply from HESSs to EVs, this paper proposes a dual-layer multi-mode (DLMM) EMOS. This strategy comprises two layers. The upper layer is a backpropagation neural network (BPNN) model enhanced by the particle swarm optimization (PSO) algorithm. It is used for real-time HESS power demand prediction. In the lower layer, a HESS operational mode determination process is formulated, and an objective optimization function is established based on HESS power loss. Under constraints designed according to the HESS state parameters, the PSO algorithm is utilized to search for the optimal power allocation ratio of the HESS in real time. The proposed DLMM-EMOS strategy is capable of providing optimal power reference values for the batteries and ultracapacitors of the HESS. The DLMM-EMOS is tested on an electrical experimental platform using US06, NEDC, and WLTP driving cycles. Results indicate that the DLMM-EMOS effectively reduces the HESS power loss while enhancing the driving range of the battery.

In a single phase, two-stage photovoltaic (PV) grid-connected system, the transient power mismatch between the dc input and ac output generates second-order ripple power (SRP). To filter out SRP, bulky electrolytic capacitors are commonly employed. However, these capacitors diminish the power density and reliability of the system. To address this issue, this paper introduces a power decoupling method. This method utilizes a bidirectional buck–boost converter, connected in parallel to the DC link, to divert SRP to a small capacitor within the single-phase grid-connected PV inverter, eliminating the need for electrolytic capacitors. The proposed topology consists of a dc–dc stage, a decoupling stage and an inverter stage, where each stage is controlled independently. In consideration of the instantaneous power fluctuations on the filtering elements, the optimal voltage reference of the decoupling capacitor is derived and implemented in the proposed decoupling control strategy. Thus, the capacitance for decoupling is minimized and the volume of the inverter is reduced. Discontinuous current control is adopted to charge and discharge the decoupling capacitor, which simplifies the decoupling control design. Finally, the steady-stage and dynamic responses of the proposed inverter are validated by simulation and experimental results in a 1-kW PV prototype.

The rotor structure of a synchronous reluctance machine (SynRM) affects the variation of reluctance and has an important effect on the torque performance of the machine. To improve the torque performance of SynRM, the optimization design of the rotor structure is essential. In this paper, a multi-objective snake optimizer (MOSO) is proposed by combining snake optimizer with multi-objective optimization strategies. Integrated with finite element analysis and the MOSO optimization method, the SynRM with a fluid-shaped rotor structure is optimized to improve torque performance. Compared with several algorithms, such as multi-objective genetic algorithm, multi-objective particle swarm optimizer, etc., the MOSO optimization method achieves better results in optimizing the fluid-shaped rotor structure with higher torque and lower torque ripple of the machines on its Pareto frontier, which verifies the superiority of the proposed MOSO optimization method. In addition, a Bezier-shaped flux barrier tip determined by curve fitting with arbitrariness is proposed to reduce the torque ripple of SynRM further. Different machines on the Pareto front of the MOSO optimization method are selected to optimize the shape of the flux barrier tip. The percentages of reduction in torque ripple for the SynRMs that have been designed with Bezier-shaped flux barrier tips are 27.27%, 47.21%, 69.71%, and 78.71%, respectively. The decrease of torque ripple for the several SynRMs verifies the effectiveness of the Bezier-shaped flux barrier tip.

The axial field flux-switching magnetic gear composite machine (AFFSMGCM) is a new type of magnetic field modulation machine with a dual-rotor. Due to the complicated structure of the AFFSMGCM and the nonlinear characteristic of dual magnetic fields coupling, a divided-layer varying-network magnetic circuit (VNMC) method is developed to optimize the machine to obtain a high calculation accuracy and reduce operation time. First, an accurate VNMC model is established according to the magnetic field distribution of an AFFSMGCM. The magnetic field coupling of the rotor and stator permanent magnets (PMs) is performed by the rotary magnetic modulation ring (RMMR). Thus, the magnetic circuit of the RMMR is divided into two layers to reduce the influence of magnetic saturation and leakage flux on the calculation accuracy of the permeances. Then the particle swarm optimization (PSO) method is used to achieve multi-objective optimization of the AFFSMGCM based on the divided-layer VNMC for achieving a large torque, low torque ripple and high efficiency. Next, a multi-physics field coupling analysis is carried out to verify the optimized AFFSMGCM. Finally, a prototype is built and experiments are carried out to validate the AFFSMGCM.

To address the issue of power utilization system redundancy in methods focusing solely on either module solar-tracking or electrical maximum power point tracking (MPPT) to enhance photovoltaic (PV) generation efficiency, the integration of PV module solar-tracking with inverter maximum power tracking is proposed to streamline the system. Combined with cost–performance analyses of single-axis and dual-axis solar-tracking, the surface characteristic relationship between the single-axis azimuth change of the module and the output voltage and output power is studied, and the output characteristic surface of the PV system is constructed. In this paper, a maximum power tracking algorithm without a position sensor on the output characteristic surface is designed, requiring only the acquisition of electrical parameters from the PV modules. Experimental comparison and analysis show that the algorithm effectively combines the azimuth tracking and the electrical maximum power tracking algorithm without a position sensor, and realizes the maximum power output of a single PV module. Under the condition of simulation settings, the output power of module is increased by 14.97–21.91%, and the daily power generation is increased by 20.42% on average under varying experimental weather conditions, providing a new idea for the effective utilization of solar energy by PV modules.

In high-voltage and high-power applications, continuous pulse-width modulation methods (CPWM) suffer from reduced inverter efficiency due to high switching frequency, and the common-mode voltage (CMV) generated can also cause electromagnetic interference and current harmonics. This paper proposes a discontinuous pulse-width modulation (DPWM)-based CMV suppression method to suppress CMV and reduce switching losses simultaneously. CMV values associated with various reference voltage vectors are analyzed first, and then CMV suppression conditions in different sectors are calculated. The proposed control strategy can correct the modulation wave of DPWM easily using a look-up table, which lists the suppression conditions calculated above to decrease the computation burden, thus suppressing CMV while simultaneously reducing switching losses. The strategy exhibits universality across various types of DPWM. Simulations and experiments confirmed its effectiveness.