With the advancements in Industry 4.0 and the global demand for sustainable energy, power electronics devices are emerging as pivotal components for modern industrial, transportation, and energy delivery. Ensuring the security of these devices is of paramount importance for safeguarding the reliable operation of critical systems against any threats. In recent years, the issue of intentional electromagnetic interference (IEMI) attack and their related countermeasures have gained widespread attention in both military and civilian power electronics sectors. Here, a review of IEMI in power electronics is provided. Starting from the electromagnetic compatibility standards, the test setups for evaluating the device immunity are introduced. Due to different attack modes and propagation paths, IEMI illuminations are typically categorized into conduction and radiation, where their susceptibility studies, applications, and defence strategies are discussed, respectively. Finally, the existing challenges, ongoing trends, and future topics in this field are discussed.

This article proposes a multiagent (MA) deep reinforcement learning (DRL) based autonomous input voltage sharing (IVS) control and triple phase shift modulation method for input-series output-parallel (ISOP) dual active bridge (DAB) converters to solve the three challenges: the uncertainties of the dc microgrid, the power balance problem, and the current stress minimization of the converter. Specifically, the control and modulation problem of the ISOP-DAB converter is formed as a Markov game with several DRL agents. Subsequently, the MA twin-delayed deep deterministic policy gradient (MA-TD3) algorithm is applied to train the DRL agents in an offline manner. After the training process, the multiple agents can provide online control decisions for the ISOP-DAB converter to balance the IVS, and minimize the current stress among different submodules. Without accurate model information, the proposed method can adaptively obtain the optimal modulation variable combinations in a stochastic and uncertain environment. Simulation and experimental results verify the effectiveness of the proposed MA-TD3-based algorithm.

The dual active bridge (DAB) converter has become a popular isolated solution to integrate energy storage systems (ESSs) and dc microgrids (MGs). However, constant power loads (CPLs) and pulsed power loads (PPLs) may reduce system damping and cause voltage oscillations in DAB converter-based ESSs (DAB-ESSs). An artificial neural network-based active disturbance rejection control (ANN-ADRC) is proposed to regulate constant output voltage quickly and accurately under different operating conditions. Firstly, the ADRC controller is designed based on the small-signal modeling of the DAB-ESSs. Feedforward compensation and uncertainty estimations of the extended state observer (ESO) help to improve the dynamic performance and to reduce the number of current sensors. Then, after satisfying the conditions of stability analysis, the parameters of the ADRC controller are selected automatically via ANN. The ANN is trained with two inputs (ADRC controller parameters) and two outputs (performance indicators of the ADRC controller). The well-trained ANN can be used as a surrogate model to obtain the optimal solution of the objective function easily and quickly. The proposed ANN-ADRC algorithm with selected parameters is implemented and validated on the hardware experimental setup. The experimental results illustrate that the proposed controller can achieve fast dynamic performance under various operating conditions.

The dc solid-state transformer (dcSST) plays a vital role in interconnecting diverse dc sources and loads in dc microgrids. However, output voltage regulation and submodule power balance control have always been two essential control challenges of the dcSST. To address these challenges, this article proposes a multiagent soft actor-critic-based active disturbance rejection control (MASAC-ADRC) method. The ADRC method is used for uncertainty estimation and compensation in modular dcSST. Specifically, the controller gains of the ADRC method are optimized adaptively by the MASAC method, thus enhancing the adaptability to changing conditions from the environment. Compared with the existing controller gain optimization methods, the proposed method does not rely on predetermined datasets. Instead, it provides a tailored strategy, employing a neural network to map optimal ADRC parameters from the measured states. Through the incorporation of diverse environmental scenarios, encompassing variant input voltage and output power, the MASAC-ADRC method achieves superior dynamic performance. Experimental validation underscores the efficacy of the proposed algorithm, showcasing its superiority over alternative approaches. The proposed method yields enhancements exceeding 50% in dynamic performance metrics such as overshoot, settling time, and mean square error when compared with existing methods.

Common-mode chokes (CMCs) are essential components of electromagnetic interference (EMI) filters. Obtaining precise common-mode (CM) and differential-mode (DM) impedance information of CMCs facilitates their performance evaluation and behavioral modeling, ultimately contributing to optimal EMI filter design. Among existing impedance extraction approaches, measurement methods are the most straightforward, not requiring CMC's detailed geometry and material properties. These methods typically use an impedance analyzer (IA) or vector network analyzer (VNA) with specific fixture adapters for measurements. However, these adapters introduce parasitics, which can affect measurement accuracy, particularly at high frequencies. Some solutions have been proposed but limited to single-phase CMCs. This article presents two impedance measurement methods for three-phase CMCs: single-port circuit de-embedding (SCD) and three-port network calibration (TNC). Both methods consider the parasitics of fixture adapters and provide unique solutions for accurate broadband measurements for the CM and DM impedances of three-phase CMCs. The SCD method has good versatility as it can be implemented using either an IA or VNA; however, it requires fixture adapter details for the de-embedding process. Conversely, the TNC method does not need these details but can only be implemented using an IA. Both methods are validated experimentally with good accuracy from 150 kHz to 120 MHz.

This letter proposes a multiagent soft actor-critic (MASAC) enabled multiagent deep reinforcement learning (MADRL) algorithm for output current sharing of the input-series output-parallel dual active bridge converter. The modular converter is partitioned into different submodules, which are treated as DRL agents of Markov games. Furthermore, all agents are executed decentralized to provide online control decisions after collaborative training. The proposed MASAC algorithm verified in a 150 V/50 V hardware prototype shows optimal dynamic performance.