Ali Arzani

Grid-following inverter (GFLI) systems equipped with LLCL filters offer superior harmonics attenuation but suffer from significant resonance issues and stability challenges under varying grid conditions. Traditional damping strategies, such as proportional-resonant (PR) controllers with active or passive damping, partially address these issues but often introduce added system complexity, increased losses, or degraded performance due to digital control delays. Moreover, these approaches typically lack guarantees of global asymptotic stability during transient disturbances. To overcome these limitations, this paper proposes a novel control strategy based on Lyapunov function-based current (LFBC) control for LLCL-filtered GFLI systems. The proposed approach ensures global asymptotic stability, robust dynamic performance, and effective resonance suppression without relying on passive loss components. The control framework integrates an active damping mechanism and a PR controller for reference signal generation, enhancing adaptability and eliminating the need for differentiation or explicit knowledge of filter parameters. Theoretical analysis using root-locus and Bode plots validates the controller's stability, and comprehensive design guidelines are provided. The experimental findings confirm that the presented LFBC controller markedly improved stability, fast dynamic response, and grid current response while minimizing total harmonic distortion.

Electric vehicle (EV) loads impact on the electric power utility distribution network (DN) can be best investigated when studying the corresponding EV charging stations (EVCS) in the DN. However, different types of data/metadata need to become available for properly modeling EVCSs for relevant power system studies. One metadata is EV users’ behavior or charging pattern. Generative AI synthetic data can be utilized to replicate EV user behavior as an alternative to metered EV data. In this paper, we first present the EVCS electrical data requirements for cyber-physical modeling and its impact on two main categories of power distribution system studies, i.e., operational planning and electromagnetic transient studies. Distribution planners can benefit by ensuring these data requirements are met prior to delving into relevant studies. We then describe the methods for generating synthetic EV data, along with sample results. Synthetic EV charging data have the potential to be utilized as input EVCS load-shapes to inform distribution grid techno-economic and offline/real-time cyber-physical simulations.

The uncontrolled $xy$ subspace greatly enhances the high harmonic components present in the stator currents. Researchers have put significant efforts to address this challenge by exploring the utilization of virtual voltage vectors (VVs) in both healthy and postfault operating conditions. However, it has been observed that the predetermined duty ratios assigned to these VVs are inadequate to effectively counterbalance intrinsic asymmetries. In order to address the above-mentioned challenges, the variable switching duty ratio-based direct torque control (DTC) was introduced in the literature for healthy operation. This article presents a VVs-based switching table-based DTC technique with variable duty-ratio for healthy as well as faulty conditions. This technique presents a distinct approach compared to conventional methods, wherein the duty ratios of the selected VVs are continuously tuned based on the $xy$ voltage commands generated by closed-loop $xy$ current controllers at each sampling interval during an open-phase fault condition. This new developed approach brings several benefits, including the effective elimination of $xy$ currents, without adverse effects on the average switching frequency or utilization of the dc link. The proposed control technique is verified by experimental evidence.

The deployment of distributed energy resources (DER) in electric power distribution networks (DN) yields significant spatiotemporal power flows and presents substantial operational challenges. Network reconfiguration (NR), via strategic control of switch states, has emerged as an effective tool to address issues such as power outages, voltage violations, increased losses, power quality issues, and a rise in associated operating costs. Based on the literature, NR enables up to 30% loss reduction, 5-15% improvement in voltage profiles, and significant gains in reliability and outage mitigation. This paper presents an application-focused review of NR strategies for operational planning in active DNs. It categorizes state-of-the-art literature based on applications, such as reliability enhancement, voltage regulation, power quality improvement, and cost reduction. This study evaluates the optimization formulations and solution techniques, – heuristic, metaheuristic, and mathematical programming approaches – by covering their strengths and limitations in NR problems. It also explores emerging directions involving machine and reinforcement learning, and quantum-inspired algorithms. The goal is to provide a clear road map for researchers and practitioners aiming to implement NR in modern electric power distribution networks.

Electric vehicles (EV) are growing more common on the roads, and there have been significant advancements in their charging methods. With the increase in EV usage, establishing an effective dc-stage fast charging control scheme is necessary to meet the high electricity demand during brief intervals. Consensus-based tracking control can be utilized to improve the efficiency of the EV battery charging process in a dc charging circuit. The controller is constructed using the EV battery state-space model and the buck-boost dc charger power deviation, subsequently applied to the dc charging circuit in MATLAB/Simulink. The proposed controller seeks to reach nearly zero power deviation in the EV battery chargers, thus boosting the efficiency of the charging operation. The findings show that the suggested controller efficiently regulates the EVs’ charging mechanism, considering both simultaneous and staggered arrival times. This approach reduces charging losses by 18.79 kW and improves the five-port dc charging circuit’s overall efficiency by 1.3% for simultaneous arrivals and by 1.27% for staggered arrivals.

Hybrid AC/DC microgrids (MGs) are comprised of AC and DC MGs interconnected through an interlinking converter (ILC). This paper presents a novel control strategy for double-stage ILCs (consisting of bidirectional DC-DC and DC-AC converters) in islanded hybrid AC/DC microgrids. To provide global power sharing among AC and DC DERs, a PI controller obtains a single relative power reference for the AC and DC DERs based on the energy content of the DC link capacitor. This relative power is then translated into references for the output voltage angular frequency at AC side of ILC and the output voltage at its DC side, in accordance with the AC and DC DERs droop characteristics. Then the DC-AC and DC-DC converters realize these references through voltage-controlled method (VCM) in their inner control loops. Accordingly, the proposed method features grid-forming capability at both AC and DC terminals of ILC, which provides it with some unique advantages. First is the improved resiliency; the system remains stable even if all DC DERs are disconnected or all AC DERs are disconnected. Secondly, stability margin and voltage transients are improved thanks to the PLL-less design and adaptation of VCM. Small-signal stability analysis and real-time simulations demonstrate the stability and dynamic performance of the proposed method not only in normal operation but also under DER disconnection. The results show that the steady-state AC and DC voltage deviations are less than 5% and the AC voltage undershoot is less than 1V.

The integration of battery energy storage systems (BESSs) within active distribution networks (ADNs) entails optimized day-ahead charge/discharge scheduling to achieve effective peak shaving.The primary objective is to reduce peak demand and mitigate power deviations caused by intermittent photovoltaic (PV) output. Quasi-static time-series (QSTS) co-simulations for determining optimal heuristic solutions at each time interval are computationally intensive, particularly for large-scale systems. To address this, a two-stage intelligent BESS scheduling approach implemented in a MATLAB–OpenDSS environment with parallel processing is proposed in this paper. In the first stage, a rule-based decision tree generates initial charge/discharge setpoints for community BESS units. These setpoints are refined in the second stage using an optimization algorithm aimed at minimizing community net load power deviations and reducing peak demand. By assigning each ADN community to a dedicated CPU core, the proposed approach utilizes parallel processing to significantly reduce the execution time. Performance evaluations on an IEEE 8500-node test feeder demonstrate that the approach enhances peak shaving while reducing QSTS co-simulation execution time, utility peak demand, distribution network losses, and point of interconnection (POI) nodal voltage deviations. In addition, the use of smart inverter functions improves BESS operations by mitigating voltage violations and active power curtailment, thereby increasing the amount of energy shaved during peak demand periods.