The application of Artificial Intelligence (AI) in the energy sector offers new opportunities for developing flexible, efficient, and sustainable infrastructures. Nevertheless, real-world deployment is still constrained by the lack of large-scale, integrated environments that can evaluate advanced algorithms under realistic operating conditions while ensuring regulatory compliance. This paper presents EnerTEF (which stands for Energy Testing and Experimentation Facility), a federated platform for testing and experimentation in the energy sector designed to address this gap. We introduce a unified TEF architecture that enables full-stack evaluation of intelligent systems, including predictive modeling, optimization, learning under data distribution shifts and federated learning across geographically distributed sites. The framework integrates high-fidelity digital twins, a privacy-preserving data exchange framework and regulatory sandboxing to support transparent, explainable and robust AI development. EnerTEF demonstrates how such a framework can be deployed in critical energy domains through three real-world scenarios including short-term hydropower generation forecasting, coordination between distribution network operators and distributed energy resources and real-time optimization of self-consumption for municipal buildings. Results show that EnerTEF effectively enables the development of novel AI models, improves cross-context generalizability and supports innovation for complex energy infrastructures, ultimately creating a practical, scalable path for addressing different energy-related problems and heterogeneous data.

The current Distribution Grid Automation (DGA) Systems are being heavily dependent on the Information and Communication Technologies (ICT) infrastructure for its proper operation. The DGA architectures are predominantly centralized and usually deployed on a dedicated hardware. This increases the risk of blackouts under a coordinated cyber-physical attack. The compromise of the dedicated hardware that hosts the central coordinator of the DGA automation results in a blackout. Though many countermeasures have already been proposed for tackling different types cyber and physical attacks on the ICT infrastructure, very few measures have been proposed to ensure the availability of the grid operation functions, even when it is compromised. This study proposes an automatic, distributed approach based on Blockchain and Smart Contract that ensures the availability of the core DGA functions even if the central coordinator that operates the grid is compromised. This is done by virtualizing and migrating/re-initialising these functions from the dedicated hardware that was compromised to another. Additionally, a Multi-Attribute Decision Making based method is incorporated into the Smart Contract that helps in selection of the optimal hardware that can host the function considering its limitations (hardware and software). Finally, a proof of concept implementation of the proposed solution is presented that utilizes the Calvin IoT (Internet of Things) platform, Flow programming tool and Hyperledger fabric and its performance is evaluated.

This paper presents experiences of laboratory testing and field demonstration of a distribution network congestion management algorithm. Development process for new smart grid functionalities and automation architectures is discussed both in general level and through the example case of congestion management. The practical implementation of the developed algorithm is explained and test setups in all development process steps (offline simulations, laboratory testing and field demonstration) are presented. The objectives in different stages of the algorithm development and testing process are discussed and lessons learned from all the steps are represented.

This paper proposes the Linear Swing Dynamics-based Virtual Synchronous Generator (LSD-VSG) for HVDC systems. The aim is to define a new role and behavior for HVDC systems in their participation to AC system frequency support. This results in enhancing the frequency stability of HVDC connected-AC grids based on the LSD concept, whereas preserving a stable operation of HVDC system. Small signal stability and parametric sensitivity analysis have been performed to support the viability of the proposed LSD-VSG and study the sensitivity of the critical eigenvalues to the closed loop system states and corresponding stability region. Additionally, various time-domain simulations are performed and the results prove the effectiveness and features of the LSD-VSG.

This paper presents the two level optimization algorithms: a centralized day-ahead and decentralized minute-ahead algorithm for energy management in an integrated mainline railway system. All energy players, such as trains, infrastructure facilities, wayside storages, and distributed energy resources, are considered in the simulation. The algorithms are developed to demonstrate the railway energy management system architecture. This paper demonstrates the validity of the algorithms and analyzes the simulation results in offline and online real case studies. In the online case study, the developed system for minute ahead optimization and real-time operation was tested on the Malaga-Fuengirola line (Spanish railway) for a few hours. The optimization is done regarding three different objectives: cost optimization or energy consumption optimization or power demand optimization.

This paper presents an implementation of an IEC 61850-based real-time co-simulation platform for verification of the performance of a volt-VAR optimization (VVO) engine for smart distribution networks. The proposed VVO engine is able to minimize grid loss, volt-VAR control asset operational costs, and conservation voltage reduction operational costs through its comprehensive objective functions, weighted by fuzzification using advanced metering infrastructure (AMI) data. The optimization engine receives the AMI data stream through measurement aggregators. Moreover, it sends control commands to volt-VAR control components modeled in real-time digital simulator (RTDS) through DNP.3 protocol. To check the performance and the precision of proposed VVO, a fault scenario is imposed upon the system. IEC 61850 GOOSE messages are generated and sent to change the status of specified breakers, while the VVO engine receives system reconfiguration commands via IEC61850 Manufacturing Message Specification (MMS) protocol. The results of the study on 33-node feeder showed adequate performance of proposed VVO in grid operating scenarios.

The concept of an all-electric ship, while offering unprecedented advantages from the point of view of efficiency and flexibility of operation, has introduced new challenges in terms of stability and power flow control. The advent of a full power electronics power system has raised new questions from the point of view of system dynamics, particularly when dealing with the new medium-voltage direct current distribution. The overall goal of guaranteeing a secure operation of the power system has brought researchers to consider two main approaches: reducing the dynamics of the large load to operate in a range of dynamics compatible with traditional generation systems, or making the generator set smarter through its power electronics interface. This paper compares these approaches to stable operation, focusing on the latter considered more in line with the progress of technology and in general more appealing.

Migration from today's power systems to future smart grids is a necessity as the energy demand continues to grow and an increasing amounts of renewable energies need to be accommodated in the grid. One of the key enablers of the smart grid is the integration of information and communication technology (ICT) into the grids in order to monitor and control power generation, distribution, and demand. Considering the close interdependence of future smart grids and communication networks, there is a need for numerical simulation to thoroughly understand the impact of the communication networks on the performance of power system dynamics, and vice versa. This paper provides an overview of available simulation techniques for smart grid communications with a focus on cosimulation frameworks and their enabling technologies. A decision tree comparing relative advantages of available cosimulation platforms and providing guidelines on how to select from them for a given application is presented. A case study analyzing agent-based shipboard smart grid protections with VPNET is presented.

Measurements validation is a critical feature in monitoring systems required by most industry applications to achieve higher level reliability. This paper presents the use of the measurement thresholds generated from the propagation of parametric uncertainty using polynomial chaos theory (PCT) to validate the sensor measurements of an AC motor drive by means of fuzzy techniques. If measurements fail the validation check, they are replaced by reconstructed data to maintain the operation. Reconstruction is performed with a PCT observer, which also supports the evaluation of the thresholds. The algorithms proposed here have been implemented and tested both in simulation and in real time experiments on a field oriented controlled induction machine.