High-frequency transformers, the core component of power electronic transformers, operates under high-frequency, square voltage. At the rising/falling edge of the squared voltage wave, overvoltage and uneven electric field distribution occur among transformer windings, especially among the fast few turns. This can cause uneven electric stress in the dielectric space and local concentration of the electric field in the insulation of the transformer, which can easily lead to premature aging of the insulation or even insulation breakdown. This paper proposes to use COMSOL and SIMULINK co-simulation to calculate the transient electric field distribution of the transformer during the rising/falling edge of the input voltage, providing a basis guideline for the insulation design of high-frequency transformers.

The utilization of DC collection and DC transmission technique, which is also called all-DC technique makes it possible to integrate large-scale wind power plant (WPP) into the grid with high efficiency and reliability. Many of these wind farms form WPP clusters. Their control system architecture is challenging and has a significant impact on the overall system performance. This paper proposes suitable control system architectures for all-DC WPP clusters, taking into account the multi-layer control framework of a large-scale system. Meanwhile, several key indicators are suggested to compare and analyze the performance of the control system architectures of both the proposed and the conventional ones in terms of response speed, investment cost, reliability, and energy collecting efficiency. The final analysis results verify that the two hierarchical control architectures are more preferable for all-DC WPP clusters under two different typical requirements.

The modular multilevel converters(MMCs) are attractive topologies for voltage sourced converter based high voltage direct current(VSC-HVDC) transmission system. Asymmetric grid conditions are common encountered in VSC-HVDC system. The resulting voltage sag and power fluctuation from such conditions amplify current stresses on devices, posing a threat to the safe and reliable operation of MMC. This article establishes an equivalent model for MMC under asymmetric conditions, based on which, a decoupling control method for the MMC under asymmetrical operating conditions is proposed.

The T-type alternate arm multilevel converter (T-AAMC) is one of the most competitive topology for high voltage DC transmission systems, which combines the lower costs and fault ride-through capability. Since the DC arms conduct alternately, the voltage operating range is narrow due to the energy balance. To expand the voltage operating range, an arm phase-shift modulation method is proposed in this paper. This paper presents an analysis of the operational process of T-AAMC to describe the relationship of the voltures and currents in DC and AC arms. Moreover, the power device, submodule capacitance and converter loss are discussed under the proposed topology and modulation. Finally, simulation results are provided to validate the effectiveness of the proposed topology and modulation strategy.

The development of large photovoltaic power stations has become a key path for achieving the carbon peak and neutrality targets. The modular structure of the three-phase Cascaded H-Bridge(CHB) grid-connected converter can easily expand its voltage level and power level, and connect to the grid without heavy transformer. In order to better support the operation of the power grid, this digest introduces the control model and design method for grid forming CHB converter based on droop control. This digest first introduces the overall structure of the CHB grid-connected converter system. The mathematical model of the three-phase CHB grid-connected inverter is derived. Based on this mathematical model, the design of the voltage current dual-closed loop and the outer power loop is introduced. The overall control structure is verified by the simulation.

Following the improvement of power converter integration, thermal system design has become an increasingly important issue. The rectangular fin heatsink is widely used in the cooling system of low-power converter because of its reliability and low cost. In this paper, the average convective heat transfer coefficient model of natural convection heatsink is given according Nusselt number experimental correlation. Furthermore, based on the proposed model, the optimal design procedure of heatsink under the natural convection is built. Finally, the effectiveness and accuracy the proposed model and optimal design procedure are verified by the CFD simulation.

The stability problems caused by converter interaction challenge the safe operation of DC microgrids. Online stability monitoring methods need to be developed urgently to prevent instability risk for DC microgrids. The existing measurement-based methods are invasive and need widely-deployed measurement units, which is difficult to realize. In this paper, a novel semi-supervised data-driven stability monitoring method based on empirical mode decomposition (EMD) and semi-generative adversarial network(SGAN) is first proposed in DC microgrids. The instability oscillation information can be extracted from complex DC bus voltage by the EMD algorithm, and the semi-supervised learning framework based on SGAN completes the training process with easily accessible unlabeled data and much less labeled data. The proposed method is non-invasive and significantly reduces reliance on labeled data, in contrast to existing methods. Subsequently, experimental tests on a hardware-in-loop(HIL) test platform confirm the effectiveness and high performance of the method, as evidenced by the construction of a DC microgrid model.

The integration of grid-connected inverters in large numbers poses challenges to the stability of the power grid. Impedance-based analysis methods are effective means to study the small-signal stability of power electronic systems. However, traditional impedance modeling approaches have limitations when dealing with black-box or grey-box models, especially for grid- connected inverters with unknown control structures and parameters. In this paper, a novel approach is proposed to address the challenges posed by black/grey-box models in grid-connected inverters. The approach is based on a neural network that incorporates physical information and utilizes measured control data. Compared to conventional neural network impedance identification methods, this approach exhibits significantly improved accuracy.

The modeling and design of the dual active bridge converter modulation strategy are essential to achieve comprehensively optimal performance in practical applications. The commonly used modeling approaches include knowledge-based and emerging data-based methods. Despite significantly reducing the required domain expertise and analytical complexity, the data-based approach heavily relies on the quality of the collected data. Therefore, this research focuses on removing abnormal data points and ensuring data quality for the data-based modeling approach. In this paper, a data-based modeling method with data pruning(MIDI) is proposed, which employs one-class support vector machine for data preprocessing before training the surrogate model. This approach has been validated with a design case and 1kW prototype hardware experiments.

The hot spot effect can cause damage to PV modules and seriously affect the safe and stable operation of PV systems. Fast and accurate detection of hot spot faults is of great importance to extend the life of PV modules and reduce power generation costs. In this paper, we propose a conditionally generated adversarial network-based hot spot detection method for PV modules, which can achieve accurate hot spot detection with small samples. Specifically, for the problem of sparse infrared image data of PV module hot spot, the hot spot dataset is expanded by a conditional generation adversarial network. The new dataset is segmented by a semantic segmentation network, which improves the problem of insufficient training of model parameters caused by the small amount of data in the original model. Through experimental validation, the proposed method optimizes the image segmentation effect and achieves accurate detection of hot spots of PV modules compared with the original model.