MASCEM - an electricity market simulator providing coalition support for virtual power players

This paper presents MASCEM - a multi-agent based electricity market simulator. MASCEM uses game theory, machine learning techniques, scenario analysis and optimization techniques to model market agents and to provide them with decision-support. This paper mainly focus on the MASCEM ability to provide the means to model and simulate Virtual Power Players (VPP). VPPs are represented as a coalition of agents, with specific characteristics and goals. The paper details some of the most important aspects considered in VPP formation and in the aggregation of new producers and includes a case study based on real data.

ANN based day-ahead spinning reserve forecast for electricity market simulation

Electricity market players operating in a liberalized environment require adequate decision support tools, allowing them to consider all the business opportunities and take strategic decisions. Ancillary services represent a good negotiation opportunity that must be considered by market players. This paper deals with short-term predication of day-ahead spinning reserve (SR) requirement that helps the ISO to make effective and timely decisions. Based on these forecasted information, market participants can use strategic bidding for day-ahead SR market. The proposed concepts and methodologies are implemented in MASCEM, a multi-agent based electricity market simulator. A case study based on California ISO (CAISO) data is included; the forecasted results are presented and compared with CAISO published forecast.

Ancillary services market clearing simulation: a comparison between deterministic and heuristic methods

Electricity market players operating in a liberalized environment requires access to an adequate decision support tool, allowing them to consider all the business opportunities and take strategic decisions. Ancillary services represent a good negotiation opportunity that must be considered by market players. For this, decision support tool must include ancillary market simulation. This paper proposes two different methods (Linear Programming and Genetic Algorithm approaches) for ancillary services dispatch. The methodologies are implemented in MASCEM, a multi-agent based electricity market simulator. A test case based on California Independent System Operator (CAISO) data concerning the dispatch of Regulation Down, Regulation Up, Spinning Reserve and Non-Spinning Reserve services is included in this paper.

Self-organization for scheduling in agile manufacturing

Agility refers to the manufacturing system ability to rapidly adapt to market and environmental changes in efficient and cost-effective ways. This paper addresses the development of self-organization methods to enhance the operations of a scheduling system, by integrating scheduling system, configuration and optimization into a single autonomic process requiring minimal manual intervention to increase productivity and effectiveness while minimizing complexity for users. We intend to conceptualize real manufacturing systems as interacting autonomous entities in order to build future Decision Support Systems (DSS) for Scheduling in agile manufacturing environments.

Personality, emotion, and mood in agent-based group decision making

Group decision making plays an important role in organizations, especially in the present-day economy that demands high-quality, yet quick decisions. Group decision-support systems (GDSSs) are interactive computer-based environments that support concerted, coordinated team efforts toward the completion of joint tasks. The need for collaborative work in organizations has led to the development of a set of general collaborative computer-supported technologies and specific GDSSs that support distributed groups (in time and space) in various domains. However, each person is unique and has different reactions to various arguments. Many times a disagreement arises because of the way we began arguing, not because of the content itself. Nevertheless, emotion, mood, and personality factors have not yet been addressed in GDSSs, despite how strongly they influence results. Our group’s previous work considered the roles that emotion and mood play in decision making. In this article, we reformulate these factors and include personality as well. Thus, this work incorporates personality, emotion, and mood in the negotiation process of an argumentbased group decision-making process. Our main goal in this work is to improve the negotiation process through argumentation using the affective characteristics of the involved participants. Each participant agent represents a group decision member. This representation lets us simulate people with different personalities. The discussion process between group members (agents) is made through the exchange of persuasive arguments. Although our multiagent architecture model4 includes two types of agents—the facilitator and the participant— this article focuses on the emotional, personality, and argumentation components of the participant agent.

Context-aware emotion-based model for group decision making

Involving groups in important management processes such as decision making has several advantages. By discussing and combining ideas, counter ideas, critical opinions, identified constraints, and alternatives, a group of individuals can test potentially better solutions, sometimes in the form of new products, services, and plans. In the past few decades, operations research, AI, and computer science have had tremendous success creating software systems that can achieve optimal solutions, even for complex problems. The only drawback is that people don’t always agree with these solutions. Sometimes this dissatisfaction is due to an incorrect parameterization of the problem. Nevertheless, the reasons people don’t like a solution might not be quantifiable, because those reasons are often based on aspects such as emotion, mood, and personality. At the same time, monolithic individual decisionsupport systems centered on optimizing solutions are being replaced by collaborative systems and group decision-support systems (GDSSs) that focus more on establishing connections between people in organizations. These systems follow a kind of social paradigm. Combining both optimization- and socialcentered approaches is a topic of current research. However, even if such a hybrid approach can be developed, it will still miss an essential point: the emotional nature of group participants in decision-making tasks. We’ve developed a context-aware emotion based model to design intelligent agents for group decision-making processes. To evaluate this model, we’ve incorporated it in an agent-based simulator called ABS4GD (Agent-Based Simulation for Group Decision), which we developed. This multiagent simulator considers emotion- and argument based factors while supporting group decision-making processes. Experiments show that agents endowed with emotional awareness achieve agreements more quickly than those without such awareness. Hence, participant agents that integrate emotional factors in their judgments can be more successful because, in exchanging arguments with other agents, they consider the emotional nature of group decision making.

Adaptive learning in agents behaviour: a framework for electricity markets simulation

Electricity markets are complex environments with very particular characteristics. A critical issue regarding these specific characteristics concerns the constant changes they are subject to. This is a result of the electricity markets’ restructuring, which was performed so that the competitiveness could be increased, but it also had exponential implications in the increase of the complexity and unpredictability in those markets scope. The constant growth in markets unpredictability resulted in an amplified need for market intervenient entities in foreseeing market behaviour. The need for understanding the market mechanisms and how the involved players’ interaction affects the outcomes of the markets, contributed to the growth of usage of simulation tools. Multi-agent based software is particularly well fitted to analyze dynamic and adaptive systems with complex interactions among its constituents, such as electricity markets. This dissertation presents ALBidS – Adaptive Learning strategic Bidding System, a multiagent system created to provide decision support to market negotiating players. This system is integrated with the MASCEM electricity market simulator, so that its advantage in supporting a market player can be tested using cases based on real markets’ data. ALBidS considers several different methodologies based on very distinct approaches, to provide alternative suggestions of which are the best actions for the supported player to perform. The approach chosen as the players’ actual action is selected by the employment of reinforcement learning algorithms, which for each different situation, simulation circumstances and context, decides which proposed action is the one with higher possibility of achieving the most success. Some of the considered approaches are supported by a mechanism that creates profiles of competitor players. These profiles are built accordingly to their observed past actions and reactions when faced with specific situations, such as success and failure. The system’s context awareness and simulation circumstances analysis, both in terms of results performance and execution time adaptation, are complementary mechanisms, which endow ALBidS with further adaptation and learning capabilities.

Application-specific modified particle swarm optimization for energy resource scheduling considering vehicle-to-grid

This paper presents a modified Particle Swarm Optimization (PSO) methodology to solve the problem of energy resources management with high penetration of distributed generation and Electric Vehicles (EVs) with gridable capability (V2G). The objective of the day-ahead scheduling problem in this work is to minimize operation costs, namely energy costs, regarding he management of these resources in the smart grid context. The modifications applied to the PSO aimed to improve its adequacy to solve the mentioned problem. The proposed Application Specific Modified Particle Swarm Optimization (ASMPSO) includes an intelligent mechanism to adjust velocity limits during the search process, as well as self-parameterization of PSO parameters making it more user-independent. It presents better robustness and convergence characteristics compared with the tested PSO variants as well as better constraint handling. This enables its use for addressing real world large-scale problems in much shorter times than the deterministic methods, providing system operators with adequate decision support and achieving efficient resource scheduling, even when a significant number of alternative scenarios should be considered. The paper includes two realistic case studies with different penetration of gridable vehicles (1000 and 2000). The proposed methodology is about 2600 times faster than Mixed-Integer Non-Linear Programming (MINLP) reference technique, reducing the time required from 25 h to 36 s for the scenario with 2000 vehicles, with about one percent of difference in the objective function cost value.

Dimensionamento de uma central de miniprodução fotovoltaica para uma exploração agrícola direcionada à indústria de laticínios

Este documento apresenta o trabalho desenvolvido no âmbito da disciplina de “Dissertação/Projeto/Estágio”, do 2º ano do Mestrado em Energias Sustentáveis. O crescente consumo energético das sociedades desenvolvidas e emergentes, associado ao consequente aumento dos custos de energia e dos danos ambientais resultantes, promove o desenvolvimento de novas formas de produção de energia, as quais têm como prioridade a sua obtenção ao menor custo possível e com reduzidos impactos ambientais. De modo a poupar os recursos naturais e reduzir a emissão com gases de efeito de estufa, é necessária a diminuição do consumo de energia produzida a partir de combustíveis fósseis. Assim, devem ser criadas alternativas para um futuro sustentável, onde as fontes renováveis de energia assumam um papel fundamental. Neste sentido, a produção de energia elétrica, através de sistemas solares fotovoltaicos, surge como uma das soluções. A presente dissertação tem como principal objetivo a realização do dimensionamento de uma central de miniprodução fotovoltaica, com ligação à rede elétrica, em uma exploração agrícola direcionada à indústria de laticínios, e o seu respetivo estudo de viabilidade económica. A exploração agrícola, que serve de objeto de estudo, está localizada na Ilha Graciosa, Açores, sendo a potência máxima a injetar na Rede Elétrica de Serviço Público, pela central de miniprodução, de 10 kW. Para o dimensionamento foi utilizado um software apropriado e reconhecido na área da produção de energia elétrica através de sistemas fotovoltaicos – o PVsyst –, compreendendo as seguintes etapas: a) definição das caraterísticas do local e do projeto; b) seleção dos módulos fotovoltaicos; c) seleção do inversor; d) definição da potência de ligação à rede elétrica da unidade de miniprodução. Posteriormente, foram estudadas diferentes hipóteses de sistemas fotovoltaicos, que se distinguem na opção de estrutura de fixação utilizada: dois sistemas fixos e dois com eixo incorporado. No estudo de viabilidade económica foram realizadas duas análises distintas a cada um dos sistemas fotovoltaicos considerados no dimensionamento, nomeadamente: uma análise em regime remuneratório bonificado e uma análise em regime remuneratório geral. Os resultados obtidos nos indicadores económicos do estudo de viabilidade económica realizado, serviram de apoio à decisão pelo sistema fotovoltaico mais favorável ao investimento. Conclui-se que o sistema fotovoltaico com inclinação adicional é a opção mais vantajosa em ambos os regimes remuneratórios analisados. Comprova-se, assim, que o sistema fotovoltaico com maior valor de produção de energia elétrica anual, que corresponde ao sistema fotovoltaico de dois eixos, não é a opção com maior rentabilidade em termos económicos, isto porque a remuneração proveniente da sua produção excedente não é suficiente para colmatar o valor do investimento mais acentuado de modo a obter indicadores económicos mais favoráveis, que os do sistema fotovoltaico com inclinação adicional. De acordo com o estudo de viabilidade económica efetuado independentemente do sistema fotovoltaico que seja adotado, é recuperado o investimento realizado, sendo a remuneração efetiva superior à que foi exigida. Assim, mesmo tendo em consideração o risco associado, comprova-se que todos os sistemas fotovoltaicos, em qualquer dos regimes remuneratórios, correspondem a investimentos rentáveis.

Adaptive learning in agents behaviour: A framework for electricity markets simulation

Electricity 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.

Metalearner based on Dynamic Neural Network for Strategic Bidding in electricity Markets

The restructuring of electricity markets, conducted to increase the competition in this sector, and decrease the electricity prices, brought with it an enormous increase in the complexity of the considered mechanisms. The electricity market became a complex and unpredictable environment, involving a large number of different entities, playing in a dynamic scene to obtain the best advantages and profits. Software tools became, therefore, essential to provide simulation and decision support capabilities, in order to potentiate the involved players’ actions. This paper presents the development of a metalearner, applied to the decision support of electricity markets’ negotiation entities. The proposed metalearner executes a dynamic artificial neural network to create its own output, taking advantage on several learning algorithms implemented in ALBidS, an adaptive learning system that provides decision support to electricity markets’ players. The proposed metalearner considers different weights for each strategy, depending on its individual quality of performance. The results of the proposed method are studied and analyzed in scenarios based on real electricity markets’ data, using MASCEM - a multi-agent electricity market simulator that simulates market players’ operation in the market.

Application-Specific Modified Particle Swarm Optimization for energy resource scheduling considering vehicle-to-grid

This paper presents a modified Particle Swarm Optimization (PSO) methodology to solve the problem of energy resources management with high penetration of distributed generation and Electric Vehicles (EVs) with gridable capability (V2G). The objective of the day-ahead scheduling problem in this work is to minimize operation costs, namely energy costs, regarding the management of these resources in the smart grid context. The modifications applied to the PSO aimed to improve its adequacy to solve the mentioned problem. The proposed Application Specific Modified Particle Swarm Optimization (ASMPSO) includes an intelligent mechanism to adjust velocity limits during the search process, as well as self-parameterization of PSO parameters making it more user-independent. It presents better robustness and convergence characteristics compared with the tested PSO variants as well as better constraint handling. This enables its use for addressing real world large-scale problems in much shorter times than the deterministic methods, providing system operators with adequate decision support and achieving efficient resource scheduling, even when a significant number of alternative scenarios should be considered. The paper includes two realistic case studies with different penetration of gridable vehicles (1000 and 2000). The proposed methodology is about 2600 times faster than Mixed-Integer Non-Linear Programming (MINLP) reference technique, reducing the time required from 25 h to 36 s for the scenario with 2000 vehicles, with about one percent of difference in the objective function cost value.

Gestão de Recursos Energéticos para Apoio à Operação dos Virtual Power Players

A liberalização dos mercados de energia elétrica e a crescente integração dos recursos energéticos distribuídos nas redes de distribuição, nomeadamente as unidades de produção distribuída, os sistemas de controlo de cargas através dos programas de demand response, os sistemas de armazenamento e os veículos elétricos, representaram uma evolução no paradigma de operação e gestão dos sistemas elétricos. Este novo paradigma de operação impõe o desenvolvimento de novas metodologias de gestão e controlo que permitam a integração de todas as novas tecnologias de forma eficiente e sustentável. O principal contributo deste trabalho reside no desenvolvimento de metodologias para a gestão de recursos energéticos no contexto de redes inteligentes, que contemplam três horizontes temporais distintos (24 horas, 1 hora e 5 minutos). As metodologias consideram os escalonamentos anteriores assim como as previsões atualizadas de forma a melhorar o desempenho total do sistema e consequentemente aumentar a rentabilidade dos agentes agregadores. As metodologias propostas foram integradas numa ferramenta de simulação, que servirá de apoio à decisão de uma entidade agregadora designada por virtual power player. Ao nível das metodologias desenvolvidas são propostos três algoritmos de gestão distintos, nomeadamente para a segunda (1 hora) e terceira fase (5 minutos) da ferramenta de gestão, diferenciados pela influência que os períodos antecedentes e seguintes têm no período em escalonamento. Outro aspeto relevante apresentado neste documento é o teste e a validação dos modelos propostos numa plataforma de simulação comercial. Para além das metodologias propostas, a aplicação permitiu validar os modelos dos equipamentos considerados, nomeadamente, ao nível das redes de distribuição e dos recursos energéticos distribuidos. Nesta dissertação são apresentados três casos de estudos, cada um com diferentes cenários referentes a cenários de operação futuros. Estes casos de estudos são importantes para verificar a viabilidade da implementação das metodologias e algoritmos propostos. Adicionalmente são apresentadas comparações das metodologias propostas relativamente aos resultados obtidos, complexidade de gestão em ambiente de simulação para as diferentes fases da ferramenta proposta e os benefícios e inconvenientes no uso da ferramenta proposta.

Desenvolvimento de uma framework de argumentação Para apoio à tomada de decisão ubíqua

Os Sistemas de Apoio à Tomada de Decisão em Grupo (SADG) surgiram com o objetivo de apoiar um conjunto de decisores no processo de tomada de decisão. Uma das abordagens mais comuns na literatura para a implementação dos SADG é a utilização de Sistemas Multi-Agente (SMA). Os SMA permitem refletir com maior transparência o contexto real, tanto na representação que cada agente faz do decisor que representa como no formato de comunicação utilizado. Com o crescimento das organizações, atualmente vive-se uma viragem no conceito de tomada de decisão. Cada vez mais, devido a questões como: o estilo de vida, os mercados globais e o tipo de tecnologias disponíveis, faz sentido falar de decisão ubíqua. Isto significa que o decisor deverá poder utilizar o sistema a partir de qualquer local, a qualquer altura e através dos mais variados tipos de dispositivos eletrónicos tais como tablets, smartphones, etc. Neste trabalho é proposto um novo modelo de argumentação, adaptado ao contexto da tomada de decisão ubíqua para ser utilizado por um SMA na resolução de problemas multi-critério. É assumido que cada agente poderá utilizar um estilo de comportamento que afeta o modo como esse agente interage com outros agentes em situações de conflito. Sendo assim, pretende-se estudar o impacto da utilização de estilos de comportamento ao longo do processo da tomada de decisão e perceber se os agentes modelados com estilos de comportamento conseguem atingir o consenso mais facilmente quando comparados com agentes que não apresentam nenhum estilo de comportamento. Pretende-se ainda estudar se o número de argumentos trocados entre os agentes é proporcional ao nível de consenso final após o processo de tomada de decisão. De forma a poder estudar as hipóteses de investigação desenvolveu-se um protótipo de um SADG, utilizando um SMA. Desenvolveu-se ainda uma framework de argumentação que foi adaptada ao protótipo desenvolvido. Os resultados obtidos permitiram validar as hipóteses definidas neste trabalho tendo-se concluído que os agentes modelados com estilos de comportamento conseguem na maioria das vezes atingir um consenso mais facilmente comparado com agentes que não apresentam nenhum estilo de comportamento e que o número de argumentos trocados entre os agentes durante o processo de tomada de decisão não é proporcional ao nível de consenso final.