ISEP – GECAD – Artigos
Permanent URI for this collection
Browse
Browsing ISEP – GECAD – Artigos by Title
Now showing 1 - 10 of 309
Results Per Page
Sort Options
- 2017 IEEE competition on modern heuristic optimizers for smart grid operation: Testbeds and resultsPublication . Lezama, Fernando; Soares, João; Vale, Zita; Rueda, Jose; Rivera, Sergio; Elrich, IstvánThis paper summarizes the two testbeds, datasets, and results of the IEEE PES Working Group on Modern Heuristic Optimization (WGMHO) 2017 Competition on Smart Grid Operation Problems. The competition is organized with the aim of closing the gap between theory and real-world applications of evolutionary computation. Testbed 1 considers stochastic OPF (Optimal Power Flow) based Active-Reactive Power Dispatch (ARPD) under uncertainty and Testbed 2 large-scale optimal scheduling of distributed energy resources. Classical optimization methods are not able to deal with the proposed optimization problems within a reasonable time, often requiring more than one day to provide the optimal solution and a significant amount of memory to perform the computation. The proposed problems can be addressed using modern heuristic optimization approaches, enabling the achievement of good solutions in much lower execution times, adequate for the envisaged decision-making processes. Results from the competition show that metaheuristics can be successfully applied in search of efficient near-optimal solutions for the Stochastic Optimal Power Flow and large-scale energy resource management problems.
- Adaptative Perturbation Patterns: Realistic Adversarial Learning for Robust Intrusion DetectionPublication . Vitorino, João; Oliveira, Nuno; Praça, IsabelAdversarial attacks pose a major threat to machine learning and to the systems that rely on it. In the cybersecurity domain, adversarial cyber-attack examples capable of evading detection are especially concerning. Nonetheless, an example generated for a domain with tabular data must be realistic within that domain. This work establishes the fundamental constraint levels required to achieve realism and introduces the Adaptative Perturbation Pattern Method (A2PM) to fulfill these constraints in a gray-box setting. A2PM relies on pattern sequences that are independently adapted to the characteristics of each class to create valid and coherent data perturbations. The proposed method was evaluated in a cybersecurity case study with two scenarios: Enterprise and Internet of Things (IoT) networks. Multilayer Perceptron (MLP) and Random Forest (RF) classifiers were created with regular and adversarial training, using the CIC-IDS2017 and IoT-23 datasets. In each scenario, targeted and untargeted attacks were performed against the classifiers, and the generated examples were compared with the original network traffic flows to assess their realism. The obtained results demonstrate that A2PM provides a scalable generation of realistic adversarial examples, which can be advantageous for both adversarial training and attacks.
- Adaptive learning in agents behaviour: A framework for electricity markets simulationPublication . Pinto, Tiago; Vale, Zita; Sousa, Tiago; Praça, Isabel; Santos, Gabriel; Morais, HugoElectricity markets are complex environments, involving a large number of different entities, playing in a dynamic scene to obtain the best advantages and profits. MASCEM (Multi-Agent System for Competitive Electricity Markets) is a multi-agent electricity market simulator that models market players and simulates their operation in the market. Market players are entities with specific characteristics and objectives, making their decisions and interacting with other players. This paper presents a methodology to provide decision support to electricity market negotiating players. This model allows integrating different strategic approaches for electricity market negotiations, and choosing the most appropriate one at each time, for each different negotiation context. This methodology is integrated in ALBidS (Adaptive Learning strategic Bidding System) – a multiagent system that provides decision support to MASCEM's negotiating agents so that they can properly achieve their goals. ALBidS uses artificial intelligence methodologies and data analysis algorithms to provide effective adaptive learning capabilities to such negotiating entities. The main contribution is provided by a methodology that combines several distinct strategies to build actions proposals, so that the best can be chosen at each time, depending on the context and simulation circumstances. The choosing process includes reinforcement learning algorithms, a mechanism for negotiating contexts analysis, a mechanism for the management of the efficiency/effectiveness balance of the system, and a mechanism for competitor players' profiles definition.
- Adaptive learning in multiagent systems for automated energy contacts negotiationPublication . Pinto, Tiago; Vale, ZitaThis paper presents the Adaptive Decision Support for Electricity Markets Negotiations (AiD-EM) system. AiD-EM is a multi-agent system that provides decision support to market players by incorporating multiple sub-(agent-based) systems, directed to the decision support of specific problems. These sub-systems make use of different artificial intelligence methodologies, such as machine learning and evolutionary computation, to enable players adaptation in the planning phase and in actual negotiations in auction-based markets and bilateral negotiations.
- Adaptive Learning Process for the Evolution of Ontology-Described Classification Model in Big Data ContextPublication . Peixoto, Rafael; Cruz, Christophe; Silva, NunoOne of the biggest challenges in Big Data is to exploit value from large volumes of variable and changing data. For this, one must focus on analyzing the data in these Big Data sources and classify the data items according to a domain model (e.g. an ontology). To automatically classify unstructured text documents according to an ontology, a hierarchical multi-label classification process called Semantic HMC was proposed. This process uses ontologies to describe the classification model. To prevent cold start and user overload, the classification process automatically learns the ontology-described classification model from a very large set of unstructured text documents. However, data is always being generated and its statistical properties can change over time. In order to learn in such environment, the classification processes must handle streams of non-stationary data to adapt the classification model. This paper proposes a new adaptive learning process to consistently adapt the ontologydescribed classification model according to a non-stationary stream of unstructured text data in Big Data context. The adaptive process is then instantiated for the specific case of of the previously proposed Semantic HMC.
- Adaptive Portfolio Optimization for Multiple Electricity Markets ParticipationPublication . Pinto, Tiago; Morais, Hugo; Sousa, Tiago M.; Sousa, Tiago; Vale, Zita; Praça, Isabel; Faia, Ricardo; Pires, Eduardo José SolteiroThe increase of distributed energy resources, mainly based on renewable sources, requires new solutions that are able to deal with this type of resources’ particular characteristics (namely, the renewable energy sources intermittent nature). The smart grid concept is increasing its consensus as the most suitable solution to facilitate the small players’ participation in electric power negotiations while improving energy efficiency. The opportunity for players’ participation in multiple energy negotiation environments (smart grid negotiation in addition to the already implemented market types, such as day-ahead spot markets, balancing markets, intraday negotiations, bilateral contracts, forward and futures negotiations, and among other) requires players to take suitable decisions on whether to, and how to participate in each market type. This paper proposes a portfolio optimization methodology, which provides the best investment profile for a market player, considering different market opportunities. The amount of power that each supported player should negotiate in each available market type in order to maximize its profits, considers the prices that are expected to be achieved in each market, in different contexts. The price forecasts are performed using artificial neural networks, providing a specific database with the expected prices in the different market types, at each time. This database is then used as input by an evolutionary particle swarm optimization process, which originates the most advantage participation portfolio for the market player. The proposed approach is tested and validated with simulations performed in multiagent simulator of competitive electricity markets, using real electricity markets data from the Iberian operator—MIBEL.
- Addressing the facilities layout design problem through constraint logic programmingPublication . Tavares, José; Ramos, Carlos; Neves, JoséOne of the most difficult problems that face researchers experimenting with complex systems in real world applications is the Facility Layout Design Problem. It relies with the design and location of production lines, machinery and equipment, inventory storage and shipping facilities. In this work it is intended to address this problem through the use of Constraint Logic Programming (CLP) technology. The use of Genetic Algorithms (GA) as optimisation technique in CLP environment is also an issue addressed. The approach aims the implementation of genetic algorithm operators following the CLP paradigm.
- Adjacent Markets Influence Over Electricity Trading—Iberian Benchmark StudyPublication . Morais, Hugo; Pinto, Tiago; Vale, ZitaThis paper presents a study on the impact of adjacent markets on the electricity market, realizing the advantages of acting in several different markets. The increased use of renewable primary sources to generate electricity and new usages of electricity such as electric mobility are contributing to a better and more rational way of living. The investment in renewable technologies for the distributed generation has been creating new opportunities for owners of such technologies. Besides the selling of electricity and related services (ancillary services) in energy markets, players can participate and negotiate in other markets, such as the carbon/CO2 market, the guarantees of origin market, or provide district heating services selling of steam and hot water among others. These market mechanisms are related to the energy market, originating a wide market strategy improving the benefits of using distributed generators. This paper describes several adjacent markets and how do they complement the electricity market. The paper also shows how the simulation of electricity and adjacent markets can be performed, using an electricity market simulator, and demonstrates, based on market simulations using real data from the Iberian market, that the participation in various complementary markets can enable power producers to obtain extra profits that are essential to cover the production costs and facilities maintenance. The findings of this paper enhance the advantages for investment on energy production based renewable sources and more efficient technologies of energy conversion
- Agent-based architecture for demand side management using real-time resources’ priorities and a deterministic optimization algorithmPublication . Gomes, Luis; Spínola, João; Vale, Zita; Corchado, Juan M.Microgrids and smart grids are largely accepted concepts in power energy systems. They bring an innovative and distributed view to the old centralized system. This demands a more active participation from end-consumers, that can be achieved by using demand response and demand side management. In this paper it is proposed a solution for demand side management involving and agent-based architectures that was deployed in a small office. The deployment integrated an algorithm for generation and consumption balance with real-time contextual resources' priorities. The deployment's overall results, from a winter and a summer day, are presented in this paper.
- Aggregation and Remuneration of Electricity Consumers and Producers for the Definition of Demand-Response ProgramsPublication . Faria, Pedro; Spínola, João; Vale, ZitaThe use of distributed generation and demand-response (DR) programs is needed for improving business models, namely concerning the remuneration of these resources in the context of smart grids. In this paper, a methodology is proposed in which a virtual power player aggregates several small-sized resources, including consumers participating in DR programs. The global operation costs resulting from the resource scheduling are minimized. After scheduling the resources in several operation scenarios, clustering tools are applied in order to obtain distinct resources’ groups. The remuneration structure that better fits the aggregator goals is then determined. Two clustering algorithms are compared: 1) hierarchical; nd 2) fuzzy c-means clustering. The remuneration of small resources and consumers that are aggregated is made considering the maximum tariff in each group. The implemented case study considers 2592 operation scenarios based on a real Portuguese distribution network with 548 distributed generators and 20 310 consumers.