In terms of low-voltage (10kW) industry applications, power converters without any flying capacitors are preferred to ensure high efficiency, high power density, and light weight. As for these requirements, this paper proposes a four-level hybrid neutral point clamped (4L-HNPC) converter and a four-level hybrid active neutral point clamped (4L-HANPC) converter. Compared to the existing three-level and four-level converters, the proposed 4L-HNPC and 4L-HANPC converters show higher efficiency or more even loss distribution. Compared to five-level converters, the proposed converters show higher power density and lower weight due to the absence of flying capacitors. To solve the DC-link capacitor imbalance issue, this paper proposes a variable- reference voltage balancing control method that aims at controlling the middle capacitor voltage at 1/3 of the DC-link voltage via a proportional integral (PI) regulator. Meanwhile, the other two capacitor voltages are automatically balanced. Compared with the existing voltage balancing control methods, the proposed variable-reference voltage balancing control method is more suitable for digital implementation as only one triangular carrier is used. Both simulation and experimental studies have been carried out to verify the proposed four-level converter topologies and variable-reference voltage balancing control method.

Hybrid switches are attracting the increasing attention recently, due to their high efficiency and low cost. Normally, two separated gate drivers and driving signals are needed for the hybrid switch, which leads to more cost and more control complexity. A novel gate driver circuit for SiC+Si hybrid switches is proposed using single driving signal and single gate driver in this paper. Two simple RC delay circuits are used for generating different time delays required by two different of triggering patterns to provide more flexibility. The detailed operation principles, mathematical analysis, and the guidance for the selection of key components of the proposed circuit are provided. The merits of the proposed driver are, first, simple structure; second, low control complexity; third, more flexibility; fourth, low cost. A 2kW SiC-MOSFET+Si-IGBT hybrid switch Buck converter is built to validate the proposed gate driver circuit. The analysis and experimental results match very well, which shows the effectiveness and merits of the proposed method.

This article considers the fault diagnosis problem for the dual active bridge (DAB) converter. Given that information from multiple diagnostic signals needs to be combined effectively for fault diagnosis of the DAB converter and diagnostic signals have noise, a multibranch 1-D convolutional neural network (CNN) based on attention mechanism is proposed. First, the proposed network processes multiple raw 1-D diagnostic signals without extracting features manually. Then, soft thresholding and attention mechanism are combined to acquire denoising thresholds, which adaptively removes channel noise from feature maps. Besides, the attention mechanism is utilized to adaptively obtain importance weights of the feature maps learned by different 1-D CNN branches, which helps to learn the most useful features and fuse the multibranch features effectively. Finally, the diagnostic results are obtained by inputting the fused features into the Softmax layer. The experimental results indicate that the proposed method with denoise module and feature fusion module improves the diagnostic accuracy effectively. Contrastive experiments show that the proposed method is superior to other fault diagnosis methods.

The hybrid switch (HyS) configured by paralleling silicon Si-IGBT and silicon carbide SiC-mosfet mosfet (SiC/Si HyS) has been widely investigated because of high efficiency as well as low cost. However, one of the main obstacles in the practical application of HyS is the complicated gate driver circuits. Generally, one HyS requires two driver chips and two driving signals to achieve flexible triggering patterns, which leads to more cost and more control complexity. To solve the problems, in this article, a compact gate driver circuit with one driver chip and one driving signal is proposed for HyS working in all commonly used triggering patterns. The operation principle, mathematical analysis of time delay, maximum switching frequency, and the relationship between delay time and triggering pattern are investigated in detail. The modeling and systematical analysis of the turning off transient process are given to clarify the current overshoot generation mechanism for the first time, which is helpful for the designer to consider the device ratio for the reliable operation of HyS. The merits of the proposed driver circuits are simple structure, low control complexity, and low cost, more importantly, providing more flexibility for realizing the commonly used triggering patterns. The feasibility and effectiveness of the proposed HyS gate driver circuit are evaluated by a 2-kW SiC/Si HyS Buck converter prototype with triggering patterns C, D, E, and F. The gate driver circuit with variable triggering patterns design guideline is outlined in this article, which helps the enterprise to integrate a new generation of driver circuit for HyS to facilitate the application of HyS as three terminals devices.

Machine learning approaches are available for the fault diagnosis issues of the dual active bridge (DAB) converter. However, influenced by noise and multi-modal diagnostic signals, the data-driven methods are still in the initial phase for certain fault diagnosis of the DAB converter. To fill this gap, a multi-modal fusion neural controlled differential equation is presented to detect and locate the DAB converter’s fault diagnosis in early phases. An improved low-rank matrix fusion method is employed to fuse the multi-modal signals (voltage and current signals) of the DAB converter. The neural controlled differential equations with sparse transformer extract significant information from the diagnostic signal, based on which fault classifier obtains the fault state of the DAB converter. The datasets are collected from a real DAB converter under single phase shift control. Contrastive experiment results demonstrate that the presented method is superior to other fault diagnosis methods.