The growing integration of PV and energy storage systems (ESSs) poses operational challenges for power distribution grids (PDGs), mainly due to generation uncertainty and voltage-profile variations. Typical receding-horizon Model Predictive Control (MPC) fails to achieve efficient control due to forecast uncertainty. Meanwhile, it is difficult to directly improve the forecast due to limited feature availability, restricted data access, and insufficient domain knowledge. In this paper, we propose a Safe Deep Reinforcement Learning (SDRL) hybrid receding-horizon MPC to manage forecast uncertainty. Using an expert model based on historical data, the SDRL agent adjusts the receding-horizon MPC output to achieve better performance that is closer to the expert. The control objective is to minimize the impact of renewable energy on the external grid while ensuring critical grid operational safety, such as voltage levels. The proposed hybrid controller is tested under various forecast uncertainties, and the results show that it can improve receding-horizon MPC performance by 65% (reducing the PCC average daily fluctuation from 92 to 32) while ensuring voltage safety. Besides, the controller is tested over various forecast update intervals and MPC horizons to demonstrate its robustness. Furthermore, a scalability test shows that while the full SDRL suffers from scaling up, the proposed hybrid controller overcomes this issue and achieves better performance, minimizing the impact of the PDG by 42–62%.

The growing demand for electric vehicles (EV) is challenging the existing grid infrastructure. In particular, higher peak loads increase the heating of the delivery transformers, thus reducing their lifetime. Therefore, efficient EV Fleet Charging Management (FCM) method to maximize the energy supply while maintaining the safe operation of the power transformer is necessary, such as its thermal loading. However, several sources of uncertainty threaten the control performance, such as the unavailability of EV information and/or unobservable states. In addition, the complexity of physical thermal loading models also makes it challenging to incorporate them into conventional optimization-based control. This paper proposes a Hybrid Safe Deep Reinforcement Learning-based FCM method by hybridizing data-driven and model-based control considering power transformer thermal loading and aging. The proposed hybrid SDRL FCM method is based on the available information from the EV fleet and the transformer and is tested on real-world EV charging records. The results show that the proposed hybrid SDRL FCM method can provide 93% energy to the EV fleet while ensuring no power and thermal overload of the power transformer and minimize aging by 75%, compared with purely data-driven methods.

The increasing penetration of inverter-based resources (IBRs) in power systems progressively erodes the rotational inertia traditionally provided by synchronous generators, threatening frequency stability. Grid-forming (GFM) converters can emulate virtual inertia and damping, but their tuning is strongly topology-dependent and non-intuitive. This article first characterizes how network topology affects key frequency-stability metrics, then proposes an optimization framework for allocating virtual inertia and damping among GFM converters in low-inertia distribution networks. The work exploits a high-order linearized state-space model that retains the full internal dynamics of synchronous generators, GFM and grid-following converters, and passive network elements. Practical frequency-stability metrics (frequency nadir, RoCoF, and active power peaks) are minimized subject to budget and feasibility constraints. The framework is validated on three network topologies (meshed, radial, and random radial). Results show that optimal allocations are strongly topology dependent: a homogeneous distribution is near-optimal in small impedant networks, whereas radial topologies require counter-intuitive asymmetric allocations that cannot be deduced analytically, motivating the formal optimization approach.

The integration of Renewable Energy Sources (RES) through Inverter-Based Resources (IBRs) is transforming power systems from synchronous machine (SM)-dominated networks to low-inertia, decentralised systems. This shift introduces stability challenges, as IBRs bring fast dynamics that disrupt traditional behaviours. Coupled with reduced inertia and strength in upstream transmission networks, distribution networks with high IBR penetration face novel dynamic interactions. This study examines these interactions, highlighting two primary stability concerns: slow dynamics, exemplified by instabilities arising from high droop gains in Grid-Forming (GFM) IBR units, and fast dynamics, illustrated by the interplay between GFM inner current loops, network dynamics, and SG stator flux dynamics. Using advanced Electromagnetic Transient (EMT) and linearised models, the study highlights how high droop gains and slower current controllers exacerbate instability. Mitigation strategies like adjusting GFM filter capacitance or line R/X ratios are proposed, offering DSOs critical insights for planning and managing high IBR integration.

The increasing penetration of Electric Vehicles (EVs) presents challenges to the distribution grid, due to more volatile power profiles and higher peak demand. One key research question is how to accommodate EVs with limited-capacity grid equipment, such as transformers and lines. However, uncertainties from the EV side and the complexity of grid equipment models challenge the performance of the control strategies implemented. Moreover, the thermal loading of the transformer is often neglected. In this work, we propose a fully model-free, safe Deep Reinforcement Learning (DRL)-based grid-to-vehicle management strategy to avoid electric and thermal overloading of the transformer and power grid constraint violation. The management strategy is based on Projection-based Constraint Policy Optimization (PCPO) and takes only the observable information from the grid and vehicles. The target is to maximize energy delivery to the EV fleet while considering safe constraints, such as transformer thermal loading, voltage magnitude limits, and line loading limits. We compared the proposed strategy with conventional DRL and other safe DRL methods and investigated its robustness against higher ambient temperatures. The results show that the proposed strategy can deliver 92 % energy and reduce violations of the grid and transformers, while the other benchmarks deliver less than 80 %. The robustness test demonstrates that the proposed strategy is effective in various temperature. Moreover, the proposed strategy can effectively reduce at most 90 % of the transformer aging incurred by the thermal stress, compared with the uncontrolled charging.

When a portion of the low-voltage distribution network, typically a village, is disconnected from the upstream network, due to a fault or maintenance, it can be re-energized temporarily by the distribution system operator. Usually, if closing of a normally open tie-switch connected to a neighboring feeder is not possible, a diesel generator is conveyed to the islanded grid portion, and used towards temporary energy supply. This solution though can be costly, environmentally detrimental or not available. In this paper, an alternative solution towards temporary energy supply of a islanded grid is presented, relying instead only on local renewable sources, here photovoltaic sources, and a small battery. This work further improves decision-making, by providing valuable information to system operator. Given consumption and generation time-series data, the continuous supply of the islanded grid is estimated, depending on several parameters such at the starting time and the demand response scheme enforced.

The medium voltage alternating current (MVAC) networks are changing with a main driver of reaching sustainable development goals. A main change is that fossil fuel-driven synchronous generators are replaced by distributed renewable inverter-based resources (IBR). As a result of this change, there will be new challenges to keep the stability of the MVAC networks. This paper focuses specifically on small signal stability challenges of MVAC networks. This is done by analyzing a simple MVAC test network with high penetration of grid following converters and a very weeak upstream grid model as a synchronous generator. This paper identifies parameters impacting the system’s stability using the tool G2ELin. The results of this study indicate that small signal stability issues can occur when the proportion of the power supplied by the grid following (GFL) converters is high and synchronous generators run low, the observed unstable mode is related to the synchronous generator. Two factors that showed to decrease the stability of the test network are, first, the distance between the synchronous generator’s location to the GFL converters and loads, and, second, natural line capacitance. To improve the stability a QV-droop control is implemented to the GFL converters; it helps the steady state voltage, but it did not improve the stability of the unstable mode. These findings underscore the need for further studies in this area to understand what parameters are important to consider to avoid small signal stability problems as more IBRs are integrated into the MVAC network.

The integration of distributed energy resources (DERs) into medium-voltage alternating current (MVAC) grids presents significant challenges, particularly in hosting capacity limitations. As decentralization efforts increase, MVAC grids become saturated, necessitating expansion to maintain stability and efficiency. This study examines two possible expansion strategies: reinforcing the existing MVAC infrastructure or transitioning to a hybrid network incorporating medium-voltage direct current (MVDC) links. A key focus is on frequency stability, specifically the impact on the rate of change of frequency (RoCoF) and nadir during primary frequency response. To address this, small-signal models are developed and validated, enabling sensitivity analysis to identify key parameters that influence RoCoF and nadir. These models are further validated through Electromagnetic transient (EMT) simulations to demonstrate their accuracy and reliability. An assessment was performed to evaluate the effectiveness of MVDC integration in improving frequency stability relative to traditional MVAC expansion.

Detailed Electromagnetic Transient (EMT) programs have become indispensable for studying power system dynamics and stability, particularly in the context of high integration of Inverter-Based Resources (IBRs). However, most commercial software integrating eigenvalue-based modal analysis tools, employ Singular Perturbation Theory hypotheses in deriving corresponding small-signal models used alongside their EMT simulations. Such hypotheses neglect fast dynamics such as the stator flux dynamics of Synchronous Machines (SMs) and network dynamics. This article presents a comprehensive small-signal modeling approach that effectively integrates these dynamics, ensuring accurate representation of the dynamics observed in EMT simulations. By employing this method, the study explores practical scenarios using benchmark networks, demonstrating how the rapid inner current control loops of Grid-Forming (GFM) converters can interact with fast network dynamics, inducing instabilities in SMs' stator flux dynamics manifested by high-frequency sub-synchronous oscillations. The findings highlight the potential of IBRs' output filter sizing and network line parameters to dampen such high-frequency oscillations driven by the fast inner current loop dynamics of GFM-based IBRs.

With the increasing share of Electrical Vehicles (EVs) in the transportation sector, the impacts on the power distribution grid are also non-neglectable. The increasing line loadings and voltage derivations due to the increasing load are the most significant impacts on the distribution grid. To manage EV fleet charging, in recent years, Deep Reinforcement Learning (DRL) has drawn more attention due to its inherent uncertainty mitigation ability. However, the DRL methods falls short when dealing with complex time series-related problems. In this paper, we proposed DRL-based EV fleet charging management in a distribution grid to mitigate the impacts of the EV fleet on the distribution grid, i.e., line loading and bus voltage derivations. The DRL agent is built with Long Short-Term Memory (LSTM) to enhance time series processing. We investigated two structures of LSTM agents and the numbers of LSTM layers. The results show that the LSTM agent is efficient to reduce EV impacts on grid, that reducing 47% current overloading and 33% voltage out-of-limit while deliver 87% energy to EV fleet.

The growing penetration of Electric Vehicles (EVs) in transportation brings challenges to power distribution systems due to uncertain usage patterns and increased peak loads. Effective EV fleet charging management strategies are needed to minimize network impacts, such as peak charging power. While existing studies have addressed uncertainties in future arrivals, they often overlook the uncertainties in user-provided inputs of current ongoing charging EVs, such as estimated departure time and energy demand. This paper analyzes the impact of these uncertainties and evaluates three management strategies: a baseline Model Predictive Control (MPC), a data-hybrid MPC, and a fully data-driven Deep Reinforcement Learning (DRL) approach. For data-hybrid MPC, we adopted a diffusion model to handle user input uncertainties and a Gaussian Mixture Model for modeling arrival/departure scenarios. Additionally, the DRL method is based on a Partially Observable Markov Decision Process (POMDP) to manage uncertainty and employs a Convolutional Neural Network (CNN) for feature extraction. Robustness tests under different user uncertainty levels show that the data hybrid MPC performs better on the baseline MPC by 20 %, while the DRL-based method achieves around 10 % improvement.

When a portion of the low-voltage grid is islanded from the main grid (due to a fault on the upstream grid or maintenance), a temporary re-energization solution is necessary. For environmental purposes and security of supply improvement, distribution system operators are also investigating alternative re-energization solutions to the currently used mobile diesel generator, such as local photovoltaic sources and batteries if available. The black start of such an inverter-based microgrid faces many challenges, especially regarding the limited short-circuit current of small-size residential PV inverters. Since telecommunication systems and load monitoring schemes might not be available during emergencies, the re-energization process needs to be designed based on the natural behavior of loads. This paper investigates how inverter stress can be reduced during the re-energization of households, while not being able to monitor the load. For that, domestic load transient measurements are conducted and assessed as a function of their impact on the stability of the considered microgrid. Then the use of a voltage ramp as a transient mitigation strategy is analyzed based on three load categories. The work is based both on detailed electromagnetic transient (EMT) models and on measurements conducted with real-life loads in an experimental facility.

In recent years, the importance of PV generation data for distribution system operations has increased. However, some behind-the-meter solar installations are still not registered with the system operator and are not necessarily monitored at a centralized level. This "hidden" generation, therefore, increases the difficulty to operate securely and efficiently the distribution grid. This paper introduces a tool dedicated to the automatic detection of such a generation. It is designed to discriminate the nodes with and without local PV generation and is aimed at high accuracy, without local measurements, thus preserving privacy and increasing security. The tool consists of a neural network coupled with a rule-based classification algorithm, which considers only a very limited volume of data (i.e., node consumption and temperature data). Open-access consumption and solar radiation data are used to feed the simulation of a 14-nodes CIGRE distribution grid used to validate the proposed approach. The implemented solution is tested across all the nodes of the selected grid. The sensitivity of the results is analyzed by the level of PV penetration and the period of observation. The tool can recognize the nodes with a new PV installation with an accuracy of up to 100%, depending on exogenous conditions.

Traditionally, photovoltaic (PV) systems have been operated using maximum power point tracking algorithms, which force the PV arrays to produce the maximum available power at all times. Nevertheless, distribution system operators are increasingly asking for flexible power point tracking (FPPT) algorithms, which allow the regulation of the PV power to a predefined reference value. FPPTs are difficult to tune and often have non-linear behavior. It complicates the modeling of PV systems for power system stability studies. This paper proposes a simplified model that reproduces the dc-side dynamics of a double-stage FPPT-controlled PV system. In addition to its simple tuning, the key advantage of the proposed model is that it can be easily translated into differential equations, which can be used in stability analyses. The proposed model is validated on a temporal simulation as well as a small-signal stability study.

This work-in-progress paper shares our experience and student feedback based on the first two years of our international research internship program at Grenoble Institute of Technology in France made available by Virginia Tech and supported through an NSF IRES grant. This program, IRES: Track I: U.s.- France Program for INverter-based And Cyber-secure Control and Communication for eLEctric power system (PINNACLE), trains and sends about 18 U.S. students (6 each year) for 8 weeks to Grenoble to engage in research and extracurricular activities, including language instruction, several industry visits, and integration with an existing international internship program, in its G2Elab. The overarching theme of students' research projects is to enable a massively inverter-based electric power system while addressing control, communication, and cybersecurity requirements and challenges. Students' exposure to these problems, especially in a European context, is expected to help them think of innovative solutions to the U.S.'s similar challenges. Additionally, G2Elab has active collaborations with several industry partners, which facilitates industry tours and field trips. This program is mutually beneficial and strengthens our existing collaboration by providing a framework for conducting research projects of common interest. We recruit nationally for this program. The program alumni are a cohort of individuals with highly desired skills for industry and graduate programs.

Inverter-based resources (IBRs) are changing the dynamics of medium-voltage distribution grids (MVDGs), leading to concerns over slow-interaction converter-driven stability (SICDS). Although researchers have proposed numerous device-level solutions for IBR stability, culminating in the promotion of grid-forming (GFM) as opposed to grid-following (GFL) inverters, a system-level analysis from the point-of-view of the distribution system operator (DSO) is still lacking. As a first step toward standardization, this article provides some guiding principles for DSOs to prevent SICDS issues in MVDGs, by performing small-signal stability analysis and selecting a set of key parameters from state-of-the-artGFLandGFMmodels. By imposing bounds for these key parameters, DSOs could manage small-signal interactions between IBRs, as exemplified in the last section of this article over a 2-IBRs study case, later scaled to a multi-inverter configuration with five inverters in a CIGRE medium voltage distribution benchmark network.

Islanded operation is a key tool to improve the reliability and operational resilience of distribution grids. To enable islanded operation with minimal hardware changes, it is convenient to operate already installed PV units in droop-controlled grid-forming mode. This paper analyzes the main adaptive strategies that allow to implement the P/ω droop principle in double-stage units interfacing intermittent energy sources. The advantages and unwanted behaviors of each strategy are illustrated through a thorough simulation-based evaluation, which leads to a set of guidelines for the design of novel improved adaptive P/ω droop strategies.

This paper presents a theoretical analysis of voltage regulation methods based on power converters that inject power on the LV side of the MV/LV transformer. Different converter connections are explored, either series, parallel or mixed. Analysis have been conducted in order to check the efficiency of the different solutions and their limitations, considering the structure complexity and the installed power. Concerning the purely reactive compensation systems, the results show that the mixed compensation structure based on both series and parallel has the greatest impact on grid voltage regulation, but that a structure allowing to exchange both active and reactive power could be even more effective. A prototype has been developed to present a simple and efficient way to do this regulation.

Virtual synchronous generators (VSGs) are one of the most relevant solutions to integrate renewable energy in weak grids and microgrids. They indeed provide inverters characteristics of rotating machines (inertia for instance) that are useful for stabilizing the system, notably in the context of the high variability of the production. Thanks to the virtual characteristics of the VSG, the virtual parameters of the emulated synchronous machine can be optimally adapted online as a function of the electric environment of the inverter. We call that inverter’s control a polymorphic VSG. The online adaptation of the critical control parameters of the VSG helps reduce the risk of deterioration of the inverter’s constituents that might be induced by harsh events (frequent in weak grids) but, more importantly, improves the robustness of the system. In this paper, four implementations of a polymorphic VSG controller are compared on a simple microgrid study case to a complete VSG model. For the test, polymorphic VSGs have to minimize frequency and voltage oscillations while withstanding short circuits, which is typically a requirement for units in this context. One of the controls is based on recurrent optimization over a prediction time horizon, and two sub-optimal ones target practical implementation in industrial inverters with limited computational power. Results show a clear reduction in incidents in the microgrid thanks to the controllers. The error reduction with the complete polymorphic VSG is up to 100% for the voltage, 32% for the currents, and 79% for the duty ratio. Those values are decreased by 30 to 50% with the sub-optimal controllers but for a reduction in the computational burden of more than 97%. Recommendations are proposed for the development of an auto-adaptive polymorphic VSG from a high technology-readiness-level perspective, i.e., targeting a compromise between error reduction and computational burden.