In recent years, several common-ground switched-capacitor transformerless (CGSC-TL) dc-ac multilevel power converters have been introduced, providing advantages such as multilevel output voltage, voltage boosting, and mitigated leakage current. However, these structures mostly suffer from drawbacks, such as limited output voltage levels (like only five levels), lack of voltage-boosting capability, and high charging current spikes of the capacitors. This article proposes a new single-stage CGSC-TL nine-level (9L) multilevel inverter (MLI) with voltage-boosting capability and limited spikes of charging current of the capacitor, designed to be employed as a single-stage power-electronics-based interface device between renewable energy sources, such as photovoltaic (PV) systems and power grid and/or load. The proposed MLI provides several merits, such as a common-ground structure that suppresses PV-to-ground leakage current associated with PV parasitic capacitances, active and reactive power support, a wide input voltage range, and higher output voltage levels (9L) compared with other structures in the same class. Comprehensive comparative analyses, as well as simulation and experimental results, are presented to verify the performance of the proposed inverter.

The improved modulation and control methods for the conventional dual active bridge topology yield significant improvements in efficiency and reduction of passive component dimensions, which help achieve high power density. This digest presents a new Symmetric Unipolar Modulation (SUM) Strategy for the DAB converter, which effectively doubles the frequency at the transformer ($2f_{s}$), whilst keeping the switching frequency of both primary and secondary side bridge semiconductor switches at fs. Unlike the conventional technique, where the bridge switching frequency and transformer frequency are the same (fs), the proposed method will help push the frequency to the next higher level to further reduce the passive component size without overstressing active and passive components. Preliminary analysis with key simulation and measurement results are presented. More measurement results will be provided in the final paper.

This paper presents the design, analysis, and experimental validation of a self-resonant transformer employing flexible multi-layer foil (FMLF) technology for a compact, high-power dense LLC resonant converter. The proposed design integrates the transformer (TX), resonant inductor (Lr), and resonant capacitor (Cr) into a single electromagnetic structure using a PQ50/50 ferrite core and trifilar foil windings on the primary side. Secondary winding employs Litz wire to reduce AC resistance and improve current distribution. Analytical modeling, analysis, and experimental testing at 400 V input and 1 kW rated power are conducted to evaluate performance. The prototype operates at 165 kHz, maintaining stable soft-switching operation. The proposed structure demonstrates a significant reduction in bypass capacitance and in passive component count, size, and copper loss compared to conventional discrete Lr-Cr-TX designs. The results confirm that FMLF-based self-resonant transformers offer a promising pathway toward higher power density and simplified design in next-generation resonant converters.

The adoption of 800V rack-level power distribution is becoming an increasingly prominent trend in modern data center infrastructure. At this voltage level, the drawback in the traditional power architecture is the multistage power conversion dropping from 800V to 48V, then to 12V, 5V, and sub-1V to power processors, along with its cumulative energy loss at each voltage step. Replacing the multistage power conversion, this paper presents a single-stage direct-to-chip power conversion topology to deliver a regulated 1V output for an 800V input. The proposed topology reduces the voltage stress on the switches upto 1/8 of the input voltage, and is capable of operating at a resonant frequency on a range of 4 times the switching frequency. This allows the use of low voltage switches operating at higher frequencies, aiding in the reduction of passive components volume. Simulation and experimental results have been presented to validate the proposed topology.

Proton exchange membrane fuel cells (PEMFCs) play a key role in hydrogen-based energy systems; however, accurate and practical modelling remains challenging due to system nonlinearities, parameter variability, and degradation effects. This paper presents a low-complexity parameter estimation methodology for a simplified PEMFC equivalent circuit model using current-switching techniques. The approach enables direct extraction of key parameters, including internal resistance and capacitance, from transient voltage responses without requiring complex optimization or large datasets. Experimental validation was conducted using 100 W and 1 kW PEMFC systems under current loading and interruption conditions. The results demonstrate good agreement between measured and simulated voltage responses, with a maximum error below 10% and typical error levels in the range of ~1.4–3%. Compared to conventional mechanistic and data-driven models, the proposed method significantly reduces computational complexity and measurement requirements while maintaining high predictive accuracy. Moreover, the combination of the simplified equivalent circuit model with current-switching-based parameter estimation offers an effective and practical tool for electrical engineers, enabling real-time monitoring, control-oriented modelling, and seamless integration with power electronic systems. The proposed approach is particularly suitable for applications in DC microgrids and digital twin-based monitoring of hydrogen energy systems.

Power consumption for electrified railway not only reflects the power supply capacity, energy supply efficiency and energy saving level, but also indicates potential defects and risks. However, due to the complexity of the traction network and the randomness of trains, there is a challenge to estimate the power consumption accurately and economically. This paper proposes a data-driven power consumption estimation algorithm of catenary and trains for electrified railway. The unit impedance of catenary is identified via a one-train condition based on the simultaneous real-time electrical data of traction substation and section post. The underdetermined issue of multi-train condition can be converted into a problem within a finite solution domain, depending on either the average speed, speed limit, or both. Considering the time dependence and the displacement continuity of trains, the power consumption can be obtained for conditions with different number of trains operating in a power supply section. Finally, the proposed method is validated by field test and simulation data from traction and power supply calculations. Taking the field test as a reference, the identified result is accurate with an error of -0.75 % compared to the conventional method. No extra monitoring devices and computing power are required in the algorithm.

LLC converters operating across wide input voltage ranges tend to deviate from their optimal efficiency point, leading to a decline in overall performance. The flying capacitor based isolated resonant converter (FCiRC) with its frequency multiplication capability, is a viable candidate for high power density converter design. It offers a high step-down conversion from 800 V input to 48 V regulated output. Nevertheless, limiting its boundaries, the existing modulation scheme does not allow for a wide-input range operation. Addressing this shortcoming, the current work proposes three additional modulation strategies allowing the converter to operate under a wide input voltage ranging from 160 V to 960 V, delivering a regulated 48 V output. The converter maintains a stable efficiency of approximately 91 % across all modulation strategies, while preserving its key benefits-such as simplified magnetic design despite wide frequency variations across modes, and lower voltage stress on the switches. Experimental validation of the proposed modulation strategy has been presented.

Traditional DC-DC converter topologies often encounter challenges such as low voltage gain, discontinuous input current, and the absence of a common ground (CG) between source and load terminals. Moreover, many of these topologies struggle to efficiently integrate power from multiple low-voltage sources, such as photovoltaic (PV), fuel cells, and battery storage, to supply high-voltage DC at a common DC bus terminal. To overcome these challenges, this paper introduces three different dual-input single-output (DISO) DC-DC converters, derived from the traditional Buck, Boost, and SEPIC topologies. After a comparative analysis, the Boost-based DISO converter is identified as the most favorable for practical applications. The proposed DISO step-up DC-DC converter, based on the conventional boost topology, is designed and analyzed for grid-tied applications, aiming to achieve high voltage gain from low-voltage PV sources. A key advantage of this converter is the establishment of a CG between the input and output ports, while maintaining a continuous input current characteristic at both input ports. The paper explores the different operating modes and performs a steady-state analysis under various input conditions. The output voltage expression of the DISO converter, incorporating both input sources, is derived through equivalent circuit analysis. Additionally, the paper discusses the voltage and peak current stresses on different devices and provides design equations for the passive components. A comprehensive performance comparison highlights the primary benefits of the proposed topology over existing DISO topologies. Simulation results, using the PSIM simulator, validate the converter's operational characteristics, and experimental results from a 350$$ \mathbf{350} $$ W prototype converter affirm the accuracy of the analysis and demonstrate its practical performance.

Electromagnetic interference (EMI) filters are essential for mitigating conducted EMI in power electronic systems. However, conventional EMI filter designs, which rely on discrete arrangements of inductors and capacitors, contribute significantly to the overall volume and weight of the system. To address this, the proposed work introduces an integrated EMI filter based on flexible multi-layer foil (FMLF) technology, which combines inductive and capacitive components into a compact structure. A key challenge in existing FMLF-based designs, particularly bifilar structures, is the formation of undesirable bypass capacitance between adjacent foil layers, which degrades EMI suppression performance, especially at high frequencies. This digest presents a novel trifilar FMLF structure that effectively minimizes the bypass capacitance, while maintaining high integration density. A hardware prototype of the proposed design has been developed and analysed using PSM1735 frequency response analyser. The results demonstrate that the trifilar structure significantly reduces bypass capacitance compared to the conventional bifilar FMLF design. Furthermore, the proposed filter achieves notable reductions in both volume and weight relative to traditional discrete-component filters, confirming its effectiveness and potential for high-density EMI filtering applications.

Cascaded H-bridge (CHB) converters are suitable candidates for numerous applications, including electrical drives, static synchronous compensators, and battery energy storage inverters. Optimal control strategies for CHB converters have attracted significant interest in recent decades due to their flexibility in including multiple control objectives and their simple design process. However, the steady-state performance of these control strategies deteriorates if the CHB converter model has parameter mismatches and/or the submodule (SM) capacitor voltage ripples are not measured. This work proposes a Kalman filter (KF) based strategy to eliminate the steady-state error and undesired low-frequency harmonic components in the CHB converter output currents. The proposed KF strategy estimates the instantaneous arm voltage harmonics representing the converter modeling errors and unaccounted disturbances. Then, these estimated voltage harmonics are used to improve the arm current predictions and obtain a compensation term for the steady-state arm voltage references to be used by the optimal control strategy. Experimental results for three different optimal control schemes are provided for a three-phase CHB converter with nine SMs to confirm the effectiveness of the proposed KF strategy.

Direct current (DC) microgrids have emerged as a promising solution for achieving resilient, efficient, and sustainable power distribution across diverse applications. Advanced control strategies are crucial to ensure reliable operation, effective energy management, and effective integration of renewable energy sources. This paper comprehensively reviews the latest developments in control techniques in DC microgrids, emphasizing their theoretical foundations, real-world applications, and performance assessments. The review covers a wide range of sophisticated control methodologies, including backstepping control, model predictive control (MPC), passivity-based control (PBC), and sliding mode control (SMC). Each approach is analyzed in detail, highlighting its underlying principles, advantages, limitations, and specific applications in DC microgrid systems. To further illustrate the practical relevance of these strategies, case studies and real-world examples are presented, demonstrating their efficacy in various operational scenarios. Additionally, the paper addresses key research challenges and identifies opportunities for future advancements in DC microgrid control strategies. This review aims to provide valuable insights for researchers and practitioners involved in the design, optimization, and operation of DC microgrids, contributing to the advancement of sustainable energy technologies in modern power systems.

The flux density in Dual Active Bridge (DAB) converters plays a crucial role in designing a high-power density converter, especially in applications requiring compact and lightweight designs. Symmetric Unipolar Modulation (SUM) not only doubles the transformer frequency to 2fsw without increasing the switching frequency (fsw) of the semiconductor devices, but it modifies the flux density in the core from triangular (conventional modulation) to trapezoidal while reducing its peak. The trapezoidal flux density in SUM significantly reduces the peak flux density requirement, thereby reducing the core size compared to traditional modulation. Simulation and experimental results validate these findings, highlighting SUM's potential for achieving higher power density in DAB converters.

Using second-life batteries (SLBs) to build battery energy storage systems (BESSs) yields substantial environmental and economic benefits. The cascaded H-Bridge (CHB) converter has emerged as an attractive candidate to integrate SLBs into the electrical grid, allowing the unbalanced power distribution among its sub-modules (SMs) with high efficiency and a low estimated cost. However, capacity differences among SLBs pose further challenges for the control system in meeting the BESS power constraints, while balancing the state-of-charge (SoC) of SLBs. This work proposes a dual-stage model predictive control (DS-MPC) strategy to balance the SoC of SLBs using a delta-connected CHB (DCHB) converter. The formulation of the proposed DS-MPC strategy is based on a discrete-time SoC dynamic model, which considers the SM modulating signals and the DCHB circulating current reference in the rotating synchronous $dq$-frame as control inputs. In this way, the proposed DS-MPC strategy obtains optimal charging and discharging currents for each SLB-SM by solving two sequential optimizations, which include maximum current ratings and converter modulation constraints, ensuring the safe operation of the BESS. Experimental results that verify the effectiveness of the proposed DS-MPC strategy are provided for a DCHB converter with nine SMs connected to Lithium-ion SLB packs of different capacities.

The transition towards a low-voltage supply in the dc power system architecture has widened the demand for high step-down dc-dc converters with enhanced efficiency and power density. A modular multilevel resonant converter being an effective candidate for this requirement, the sizing of passive components continues to be a challenge while aiming for higher power density. Identifying the impacting variables, this paper proposes a unique modulation technique for a flying capacitor based isolated resonant converter. This technique achieves a voltage gain of 16:1 with 1:1 isolation transformer and reduces the voltage stress on the switches by quarter of the dc-link voltage. Also, the switching frequency is quadrupled, allowing the converter to operate at a resonant frequency in a range of 4 times the switching frequency. This combination caters a drastic reduction in the switching and conduction losses of the semiconductor as well as reduction in size of the passive components, contributing to high efficiency operation and push the power density to a next higher level. Simulation and experimental results are presented to validate the proposed technique.

Optimal control strategies for cascaded H-Bridge (CHB) converters have attracted significant interest in recent decades due to their high performance and flexibility in including multiple control objectives. Nevertheless, the steady-state performance of these control strategies can deteriorate if the converter model has parameter mismatches and/or sub-module capacitor voltage ripples are not measured. This work proposes a linear quadratic Gaussian (LQG) current control strategy for a CHB-based second-life battery energy storage system. The main novelty of the proposed LQG strategy lies in an augmented state space model that allows a Kalman Filter observer to estimate the instantaneous arm voltage disturbances representing the converter modelling errors in the second-life battery packs. These estimated voltage disturbances are used to enhance the arm current predictions and compute the steady-state arm voltage references used by a linear quadratic regulator current control strategy. As a result, the proposed LQG strategy eliminates the steady-state error in the CHB converter currents even if the second-life batteries are modelled as ideal constant voltage sources. Experimental results are provided for a three-phase CHB converter with nine SMs directly connected to second-life battery packs to validate the effectiveness of the proposed LQG strategy.

This article proposes a novel three-port converter with a fault-tolerant (FT) capability to reconfigure automatically under different switching fault conditions, i.e., open circuit fault (OCF) and short circuit fault (SCF) of the power transistors, and to continue achieving different control objectives such as effective battery charging, maximum power point tracking (MPPT), and output voltage regulation. The proposed fault-tolerant design with a one-level redundancy structure enhances the reliability of traditional three-port converters (TPCs). Experimental results of a laboratory prototype has proven the effectiveness of the proposed converter.

Cascaded H-bridge (CHB) converters are an attractive candidate for numerous applications, including static synchronous compensators and next-generation photovoltaic and battery energy storage inverters. Due to its simplicity, scalability, and excellent harmonic performance, phase-shifted pulsewidth modulation (PS-PWM) is one of the preferred modulation strategies for CHB converters. However, the latter advantage might be drastically affected when an unbalanced operation in the H-bridge cells is required, e.g., setting different dc-voltage levels and/or ac-voltage references among cells. This work proposes a predictive optimal variable angle PS-PWM (OVA-PS-PWM) strategy for CHB converters. The proposed OVA-PS-PWM introduces a bilinear dynamic model that describes the impact of the phase-shift angles (PS-angles) over the CHB output voltage harmonics. This model is then employed to formulate an optimal control problem that minimizes the output voltage harmonic distortion. An analytical optimal solution for a PS-angle update rule that applies to CHB converters of any number of cells is derived. As a result, the proposed OVA-PS-PWM updates each PS-angle at every sampling instant, significantly improving the harmonic content of the CHB output voltage even under severely unbalanced operation scenarios. Experimental results are provided with a CHB converter with nine cells to verify the effectiveness of the proposed optimal modulation strategy.

Increasing the penetration of PV systems into the grid can lead to several instability issues within the power system since they typically operate at the maximum power point and have a limited control margin against system dynamics. To increase the penetration of the PV system into the grid, this work proposes an innovative method that connects the droop control of the grid-forming inverter with the power controller of the PV module. The power controller delivers the required power based on droop loop calculations to emulate the conventional synchronous generators' behaviour. As long as the desired power of the droop loop is lower than the maximum power, the power controller of the PV system delivers the target power. Otherwise, it delivers the maximum power. Additionally, for grid-following inverter applications, the developed power controller effectively maintains the demand-supply balance by generating the same reference power assigned to the grid-following inverter without injecting excessive power into the grid.

An accurate state of charge (SOC) estimation is important for ensuring the safe operation of electric vehicles (EVs). Recurrent Neural Networks (RNNs), known for their time-sequence capabilities, offer advantages in battery SOC estimation. Bidirectional RNNs (BiRNN), in particular, have recently been introduced in this domain. Conventional testing methods for BiRNN often utilize the entire dataset, allowing the model to access future data for present SOC estimation, which is impractical for real-world applications. To address this issue, this paper proposes a real-time testing method that restricts the BiRNN to historical data by employing a sliding window. For comparison, the BiRNN is also tested with a same-sized sliding window containing both historical and future data. Furthermore, this paper provides a comparative evaluation of unidirectional RNNs and BiRNNs for SOC estimation. The results reveal that the performance of BiRNNs under the historical data scenario is similar to unidirectional RNNs, with a minor improvement of an average 0.08% mean absolute error (MAE) on the testing dataset. When exposed to future data, BiRNNs demonstrate an average MAE improvement of 0.25%.