Abstract
EUniversal project aims to facilitate the use of flexibility services and interlink distribution system's active management with electricity markets. Implementing market-based flexibility services implies a change in distribution network monitoring and control towards a more predictive approach. However, integrating cost-effective monitoring and control tools for the LV network is still quite challenging. Within the project, a set of operation and planning tools have been developed for a coordinated quantification and activation of flexibility in HV, MV and LV distribution networks. The paper presents the tools developed for the Portuguese pilot and shows preliminary results obtained when considering network operation scenarios characterized by large scale integration of DER and EV.

Abstract
The connection of distributed energy resources in distribution system have been increasing significantly, requiring new approaches as market-based flexibility solutions. This paper proposes the coordinated operation of on-load tap changer and flexibility services traded in a local market for voltage regulation in medium and low voltage grid. The wider action of on-load tap changer is used to restore voltages at the medium voltage feeder based on sensitivity coefficients. If voltage violations persist, flexibilities are traded in a local energy market with a cost-effective approach, where flexibility costs are minimized, and are activated according to their effectiveness indicated by sensitivity coefficients. Sensitivity coefficients are obtained in the medium voltage using an analytical approach that can be applied to multi-phase unbalanced systems, and in the low voltage using a data-driven approach due to their limited observability. Results show the proposed approach can be an effective solution to regulate voltages, combining the wider action of on-load tap changer with local flexibility, avoiding unnecessary tap changes and requesting a small volume of flexibility services.

Abstract
The growing integration of renewable energy sources and the widespread electrification of the energy demand have significantly reduced the capacity margin of the electrical grid. This demands a more flexible approach to grid operation, for instance, combining real-time topology optimization and redispatching. Traditional expert-driven decision-making rules may become insufficient to manage the increasing complexity of real-time grid operations and derive remedial actions under the N-1 contingency. This work proposes a novel hybrid AI framework for power grid topology control that integrates genetic network programming (GNP), reinforcement learning, and decision trees. A new variant of GNP is introduced that is capable of evolving the decision-making rules by learning from data in a reinforcement learning framework. The graph-based evolutionary structure of GNP and decision trees enables transparent, traceable reasoning. The proposed method outperforms both a baseline expert system and a state-of-the-art deep reinforcement learning agent on the IEEE 118-bus system, achieving up to an 28% improvement in a key performance metric used in the Learning to Run a Power Network (L2RPN) competition.

Abstract
Electric vehicles (EVs) are crucial in achieving a low-carbon transportation sector and can inherently offer demand-side flexibility by responding to price signals and incentives, yet real-world strategies to influence charging behavior remain limited. This paper combines bilevel optimization and causal machine learning as complementary tools to design and evaluate dynamic incentive schemes as part of a pilot project using a supermarket's EV charging station network. The bilevel model determines discount levels, while double machine learning quantifies the causal impact of these incentives on charging demand. The results indicate a marginal increase of 1.16 kW in charging demand for each one-percentage-point increase in discount. User response varies by hour and weekday, revealing treatment effect heterogeneity, insights that can inform business decision-making. While the two methods are applied independently, their combined use provides a framework for connecting optimization-based incentive design with data-driven causal evaluation. By isolating the impact of incentives from other drivers, the study sheds light on the potential of incentives to enhance demand-side flexibility in the electric mobility ecosystem.

Abstract
The European Union's Artificial Intelligence (AI) Act defines robustness, resilience, and security requirements for high-risk sectors but lacks detailed methodologies for assessment. This paper introduces a novel framework for quantitatively evaluating the robustness and resilience of reinforcement learning agents in congestion management. Using the AI-friendly digital environment Grid2Op, perturbation agents simulate natural and adversarial disruptions by perturbing the input of AI systems without altering the actual state of the environment, enabling the assessment of AI performance under various scenarios. Robustness is measured through stability and reward impact metrics, while resilience quantifies recovery from performance degradation. The results demonstrate the framework's effectiveness in identifying vulnerabilities and improving AI robustness and resilience for critical applications.

Abstract
A fast response to faults in large-scale photovoltaic power plants (PVPPs), which can occur on hundreds of components like photovoltaic panels and inverters, is fundamental for maximizing energy generation and reliable system operation. This work proposes using a Graph Neural Network (GNN) combined with a digital twin for synthetic fault data scenario generation for fault location in PVPPs. It shows that GNN can adapt to system changes without requiring model retraining, thus offering a scalable solution for the real operating PVPPs, where some parts of the system may be disconnected for maintenance. The results for a real PVPP show the GNN outperforms baseline models, especially in larger topologies, achieving up to twice the accuracy in a fault location task. The GNN's adaptability to topology changes was tested on the simulated reconfigured systems. A decrease in performance was observed, and its value depends on the complexity of the original training topology. It can be mitigated by using several system reconfigurations in the training set.