The railway power conditioner (RPC) based on the back-to-back modular multilevel converter (MMC), named multilevel hybrid RPC (MHRPC), was proposed to compensate for the power quality problems in traction power supply system (TPSS), such as negative sequence current, reactive power, and harmonics. Since TPSS is a 25 kV medium-voltage system, the phase-shifted pulsewidth modulation (PS-PWM) was chosen for MHRPC considering the number of submodules. However, the conventional PS-PWM suffers from higher distortion in output voltage or extra harmonics in circulating currents when applying to MHRPC. Various approaches were investigated to suppress harmonics, such as modified PWM methods, tailor-made filters, or additional transformers, but all increase control complexity or initial cost. In this article, a 2-D fast Fourier transform (FFT) harmonic model is developed to analyze how the phase displacement between stacks and legs impacts the harmonic distortion. Then an interpolated PS-PWM (IPS-PWM) is proposed, which sets the initial phase of carriers by coordinating four stacks of a single-phase MMC. It provides more output voltage stairs and eliminates the voltage unbalance states between the dc bus and two legs. Thus, MHRPC achieves better performance without extra hardware or increasing switching frequency. Simulation and experiment results verify the effectiveness of the proposed method.

The modular multilevel converter (MMC) has been a crucial component for medium/high-power energy conversion systems due to its modular structure. In the literature, the harmonic analysis of MMC with different modulation methods, including the conventional carrier phase-shifted pulsewidth modulation (CPS-PWM) method, have been extensively investigated. However, the influence of the carrier displacement angles between three-phase legs on the output performance of an MMC is normally ignored. This article proposes a dual-phase-shifted PWM (DPS-PWM) for a three-phase MMC, which further enhances the performance of the conventional CPS-PWM scheme through the proper selection of carrier displacement angles between three phases. First, the operation principle and implementation of the proposed DPS-PWM method are described. Then a two-dimensional Fourier model is derived to analyze how the carrier displacement angles between stacks and legs affect the harmonics of output voltage and circulating current. It is found that DPS-PWM can suppress specific harmonics in the line-to-line voltage by coordinating the carriers' phase shifting between the three-phase legs without affecting phase voltage and circulating current. Finally, the effectiveness of the DPS-PWM method is verified by simulation and experimental results.

Input voltage sharing (IVS) is necessary for the stable operation of an input-series-output-parallel (ISOP) dual-active-bridge (DAB) converter. However, the nonidentical high-frequency link (HFL) inductor is unavoidable, causing mutual interferences among submodules, and deteriorating the decoupling and voltage balance performances of conventional IVS strategies. Thus, a general decoupling IVS control strategy is proposed for ISOP DAB converter with nonidentical HFL inductors. First, an improved small-signal state-space model of ISOP DAB converter with a full-rank matrix is proposed to present the coupled performance of the ISOP DAB converter with nonidentical HFL inductors. Based on the improved model, a general decoupling method is proposed by invertible and gain-regulated matrices to achieve a diagonal matrix with uniform diagonal elements in the small-signal model, revealing and solving the sneak coupled performances among control loops. Next, a decoupling IVS control strategy is proposed by shaping full-order input impedance to solve the issues and coupled effects of multiple modes in the operation submodules. Then, the IVS controller is designed and detailed in the frequency and time domains, achieving input voltage balance and avoiding oscillation among submodules with better dynamics and more stable responsiveness. Moreover, the better characteristics of input impedance, output impedance, and input voltages in the proposed decoupling IVS control strategy are discussed and compared with traditional strategies. Finally, experiments are set up to verify the proposed general decoupling IVS control strategy for ISOP DAB converter with nonidentical HFL inductors.

The decentralized control strategy is beneficial to the expansibility of the input-series-output-parallel (ISOP) dual active bridge (DAB) converter. Under the previous decentralized control strategies, coupling disturbances might lead to voltage oscillations in the ISOP DAB converter. Besides, the traditional decoupling input voltage sharing (IVS) strategy needs a center controller, limiting the expansibility of the ISOP converters. This paper proposes a linear active disturbance rejection control (LADRC) decoupling strategy to achieve decentralized and decoupling control for the ISOP DAB converter. The coupling disturbances of the IVS and output voltage regulation (OVR) control loops are separated out, observed by the designed linear extended state observers (LESO), and restrained by the LADRC controllers. The IVS controller design considers the loop-gain ratio to maintain the positive input impedance with various-order OVR controllers. In the closed-loop analysis, the coupling disturbances are eliminated, and the IVS and OVR control loops are decoupled. Meanwhile, the stability and high-frequency ripple suppression of the ISOP DAB converter is improved. Finally, experimental results based on a dual-module ISOP DAB converter are presented to verify the advantages of the proposed strategy.

Ensuring that electrical workers use personal protective equipment (PPE) correctly is critical to electrical safety, but existing detection methods face significant limitations when applied in the electrical industry. This paper introduces MRC-DETR (Multi-Scale Re-calibration Detection Transformer), a novel framework for detecting Power Engineering Personal Protective Equipment (PEPPE) in complex electrical operating environments. Our method introduces two technical innovations: a Multi-Scale Enhanced Boundary Attention (MEBA) module, which significantly improves the detection of small and occluded targets through optimized feature representation, and a knowledge distillation strategy that enables efficient deployment on edge devices. We further contribute a dedicated PEPPE dataset to address the lack of domain-specific training data. Experimental results demonstrate superior performance compared to existing methods, particularly in challenging power industry scenarios.