Ricardo J. Bessa’s research while affiliated with Institute for Systems and Computer Engineering, Technology and Science (INESC TEC) and other places

What is this page?


This page lists works of an author who doesn't have a ResearchGate profile or hasn't added the works to their profile yet. It is automatically generated from public (personal) data to further our legitimate goal of comprehensive and accurate scientific recordkeeping. If you are this author and want this page removed, please let us know.

Publications (149)


A Conceptual Framework for AI-based Decision Systems in Critical Infrastructures
  • Preprint
  • File available

April 2025

·

27 Reads

Milad Leyli-abadi

·

Ricardo J. Bessa

·

·

[...]

·

Toni Waefler

The interaction between humans and AI in safety-critical systems presents a unique set of challenges that remain partially addressed by existing frameworks. These challenges stem from the complex interplay of requirements for transparency, trust, and explainability, coupled with the necessity for robust and safe decision-making. A framework that holistically integrates human and AI capabilities while addressing these concerns is notably required, bridging the critical gaps in designing, deploying, and maintaining safe and effective systems. This paper proposes a holistic conceptual framework for critical infrastructures by adopting an interdisciplinary approach. It integrates traditionally distinct fields such as mathematics, decision theory, computer science, philosophy, psychology, and cognitive engineering and draws on specialized engineering domains, particularly energy, mobility, and aeronautics. The flexibility in its adoption is also demonstrated through its instantiation on an already existing framework.

Download

Fig. 5. Example of cosine similarity to unperturbed state in each step between steps 350 and 750 of an episode in the environment with the RLPA
On the Definition of Robustness and Resilience of AI Agents for Real-time Congestion Management

April 2025

·

12 Reads

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.


Budget-Constrained Collaborative Renewable Energy Forecasting Market

April 2025

·

7 Reads

IEEE Transactions on Sustainable Energy

Accurate power forecasting from renewable energy sources (RES) is crucial for integrating additional RES capacity into the power system and realizing sustainability goals. This work emphasizes the importance of integrating decentralized spatio-temporal data into forecasting models. However, decentralized data ownership presents a critical obstacle to the success of such spatio-temporal models, and incentive mechanisms to foster data-sharing need to be considered. The main contributions are a) a comparative analysis of the forecasting models, advocating for efficient and interpretable spline LASSO regression models, and b) a bidding mechanism within the data/analytics market to ensure fair compensation for data providers and enable both buyers and sellers to express their data price requirements. Furthermore, an incentive mechanism for time series forecasting is proposed, effectively incorporating price constraints and preventing redundant feature allocation. Results show significant accuracy improvements and potential monetary gains for data sellers. For wind power data, an average root mean squared error improvement of over 10% was achieved by comparing forecasts generated by the proposal with locally generated ones.



Budget-constrained Collaborative Renewable Energy Forecasting Market

January 2025

·

15 Reads

Accurate power forecasting from renewable energy sources (RES) is crucial for integrating additional RES capacity into the power system and realizing sustainability goals. This work emphasizes the importance of integrating decentralized spatio-temporal data into forecasting models. However, decentralized data ownership presents a critical obstacle to the success of such spatio-temporal models, and incentive mechanisms to foster data-sharing need to be considered. The main contributions are a) a comparative analysis of the forecasting models, advocating for efficient and interpretable spline LASSO regression models, and b) a bidding mechanism within the data/analytics market to ensure fair compensation for data providers and enable both buyers and sellers to express their data price requirements. Furthermore, an incentive mechanism for time series forecasting is proposed, effectively incorporating price constraints and preventing redundant feature allocation. Results show significant accuracy improvements and potential monetary gains for data sellers. For wind power data, an average root mean squared error improvement of over 10% was achieved by comparing forecasts generated by the proposal with locally generated ones.



Enhancing the European power system resilience with a recommendation system for voluntary demand response

December 2024

·

14 Reads

iScience

Climate change, geopolitical tensions, and decarbonization targets are bringing the resilience of the European electric power system to the forefront of discussion. Among various regulatory and technological solutions, voluntary demand response can help balance generation and demand during periods of energy scarcity or renewable energy generation surplus. This work presents an open data service called Interoperable Recommender that leverages publicly accessible data to calculate a country-specific operational balancing risk, providing actionable recommendations to empower citizens toward adaptive energy consumption, considering interconnections and local grid constraints. Using semantic interoperability, it enables third-party services to enhance energy management and customize applications to consumers. Real-world pilots in Portugal, Greece, and Croatia with over 300 consumers demonstrated the effectiveness of providing signals across diverse contexts. For instance, in Portugal, 7% of the hours included actionable recommendations, and metering data revealed a consumption decrease of 4% during periods when consumers were requested to lower consumption.




AI to Enhance Power Systems: Modeling, Operation, and Control [Guest Editorial]

November 2024

·

1 Read

IEEE Power and Energy Magazine

The power system worldwide is currently undergoing a rapid and transformative shift driven by decarbonization and digitization. The global goal of decarbonization in response to climate change has driven the evolution of smart grids toward high, or even 100%, renewable penetration. Meanwhile, traditional electricity consumers are transforming into prosumers who simultaneously consume and produce electricity. More flexible resource types continue to emerge and rise on the demand side. They have emerged as highly effective solutions to integrate and coordinate a diverse range of distributed energy resources that offer increased resilience, flexibility, and sustainability during the decarbonization of power systems.


Citations (48)


... By retaining early layers of the deep neural network (DNN) to identify general features, later layers are fine-tuned for new datasets, facilitating adaptation to different system configurations. 4. Foundational models: Recent development of foundational models (e.g., large-scale pre-trained models) provide an alternative perspective, where broad generalization capabilities are achieved through training on extensive and diverse datasets [44]. Incorporating this methodology into DRL Agent training could further enhance scalability and generalization in future work. ...

Reference:

Efficient Deep Reinforcement Learning-Based Supplementary Damping Control with a Coordinated RMS Training and EMT Testing Scheme (early access)
Foundation models for the electric power grid
  • Citing Article
  • December 2024

Joule

... Two types of energy allocation modes are supported by RECreation, granting this platform the ability to adapt to different regulatory frameworks, namely pre-delivery and post-delivery allocation [94]. Pre-delivery allocation methods rely on allocation coefficients (ACs) independent of the final metered energy. ...

Decarbonized and Inclusive Energy: A Two-Fold Strategy for Renewable Energy Communities
  • Citing Article
  • July 2024

IEEE Power and Energy Magazine

... Fu et al. introduced a Copula-based joint probability distribution method for wind speed and rainfall intensity, analyzing the impact of wind-rain combined loads on the failure probability of transmission lines [23]. Kazemi-Robati et al. proposed a scenario generation method based on Copula theory to account for the dependence structure among different random variables, addressing uncertainties in renewable energy generation and energy prices [24]. ...

Stochastic optimization framework for hybridization of existing offshore wind farms with wave energy and floating photovoltaic systems
  • Citing Article
  • April 2024

Journal of Cleaner Production

... In the first step, a variety of testing initiatives, approaches, applications, and use cases from the targeted domain (i.e., power and energy systems) is gathered. Mainly project-related resources are being used for this step [6]- [8]. In the second step (cf. ...

Blueprint of the Common European Energy Data Space - Version 1.0
  • Citing Technical Report
  • March 2024

... Dozens more will be required; overall, apart from being used as feedstock for hydrogen production, some is used as a cooling fluid [32]. Overall, the green hydrogen process is beneficial in that it can sustain the intermittent nature of RE sources [37]. ...

The Role of Batteries in Maximizing Green Hydrogen Production with Power Flow Tracing
  • Citing Conference Paper
  • January 2024

... However, the substantial computational power provided by these centers is associated with significant environmental impacts [1], as they are among the significant consumers of energy and emitters of carbon emissions [2]. This has led to increasing scrutiny regarding their sustainability [3]. In response, liquid immersion cooling has emerged as a superior alternative to conventional air cooling systems, offering enhanced thermal management efficiency and ensuring the operational stability of supercomputing systems [4]. ...

A review on the decarbonization of high-performance computing centers

Renewable and Sustainable Energy Reviews

... The Joint Research Centre has developed a smart-grid-based cost-benefit analysis approach, which is further detailed in [38]. A multi-criteria cost-benefit methodology for all stakeholders interested in evaluating various planning options is outlined in [39]. ...

The benefits of smart4RES predictive analytics
  • Citing Conference Paper
  • July 2023

IET Conference Proceedings

... Knowledge-based approaches leverage historical data and machine learning techniques for fault detection [17,18]. Methods such as Principal Component Analysis (PCA) [19], Arti-ficial Neural Networks (ANNs) [20,21], and Support Vector Machines (SVMs) [22] have been used to classify fault conditions based on training data. ...

PV Inverter Fault Classification using Machine Learning and Clarke Transformation
  • Citing Conference Paper
  • June 2023

... Moreover, it facilitates and promotes new activities such as pairing SPs with potential clients via interoperable protocols, enable various flexibility service providers (FSPs) to interact with different flexibility market platforms by providing interoperable interfaces to these platforms, enables flexibility activation and settlement of SOs, and support new BMs related to consumer data processing by properly marketing this data to software companies, unlocking additional value to consumers engaged in flexibility provision. These potential BMs were already identified in [11] as a preliminary step to build the FCVC. ...

Analysis of Flexibility-centric Energy and Cross-sector Business Models
  • Citing Conference Paper
  • June 2023

... A multi-level preventive management framework was developed to enable DSO procurement of day-ahead market-based flexibility services for congestion and voltage control [1]. The main characteristics are summarized below: ...

Euniversal's smart grid solutions for the coordinated operation & planning of MV and LV networks with high EV integration
  • Citing Conference Paper
  • June 2022

IET Conference Proceedings