Abstract
Recent commitments and consequent advances towards an effective energy transition are resulting in promising solutions but also bringing out significant new challenges. Models for energy management at the building and microgrid level are benefiting from new findings in distinct areas such as the internet of things or machine learning. However, the interaction and complementarity between such physical and virtual environments need to be validated and enhanced through dedicated platforms. This chapter presents the Multi-Agent based Real-Time Infrastructure for Energy (MARTINE), which provides a platform that enables a combined assessment of multiple components, including physical components of buildings and microgrids, emulation capabilities, multi-agent and real-time simulation, and intelligent decision support models and services based on machine learning approaches. Besides enabling the study and management of energy resources considering both the physical and virtual layers, MARTINE also provides the means for a continuous improvement of the synergies between the Internet of Things and machine learning solutions.

Abstract
Complex optimization problems are often associated to large search spaces and consequent prohibitive execution times in finding the optimal results. This is especially relevant when dealing with dynamic real problems, such as those in the field of power and energy systems. Solving this type of problems requires new models that are able to find near-optimal solutions in acceptable times, such as metaheuristic optimization algorithms. The performance of these algorithms is, however, hugely dependent on their correct tuning, including their configuration and parametrization. This is an arduous task, usually done through exhaustive experimentation. This paper contributes to overcome this challenge by proposing the application of sequential model algorithm configuration using Bayesian optimization with Gaussian process and Monte Carlo Markov Chain for the automatic configuration of a genetic algorithm. Results from the application of this model to an electricity market participation optimization problem show that the genetic algorithm automatic configuration enables identifying the ideal tuning of the model, reaching better results when compared to a manual configuration, in similar execution times.

Abstract

Abstract
The large-scale integration of electric vehicles (EVs) can contribute to the better use of renewable resources and the emergence of new technologies. However, if not properly controlled, it has several downsides. Several strategies make it possible to perform this control by making use of data mining models to deal with the large amounts of data associated with EVs that need to be considered. Accordingly, this chapter presents a study on the progress of EVs integration, where the economic and socio-demographic aspects and the development of the EVs global market are highlighted. Furthermore, some recommendations are suggested to policymakers related to EV management and possibilities for future improvement of EV integration. Finally, this chapter provides a review of data mining models and applications that deal, directly or indirectly, with EV-related problems. © 2023 The Institute of Electrical and Electronics Engineers, Inc.

Abstract
Data mining approaches are increasingly important to enable dealing with the constantly rising challenges in power and energy systems. Classification models, in particular, are suitable for predicting classes of new observations based on previous cases. This chapter illustrates the advantages of the use of classification models, namely artificial neural networks and support vector machines, to predict the behavior profiles of electricity market negotiation players. A clustering model is used to identify similarities in the behavior of players, resulting in a set of negotiation profiles. The negotiation behavior of new players is then classified as belonging to one of these profiles, allowing for an automated adaptation of the negotiation process according to the expected reactions of the opponent. © 2023 The Institute of Electrical and Electronics Engineers, Inc.

Abstract
The rapid developments in Internet-of-Things (IoT), cloud computing, and big data technologies have increased the popularity of machine learning (ML) techniques. As a result, of all ML techniques, deep learning (DL) is at the forefront of innovation, outperforming all other techniques in many application domains. DL has made breakthroughs in speech recognition, image processing, forecasting, natural language processing, fault detection, power disturbance classification, energy trading, and much more. DL is a complex ML approach composed of multiple processing layers, which allows pattern and structure recognition on huge datasets. This chapter takes an in-depth look at the most recent and promising DL works in the literature for intelligent power and energy systems (PES). Several types of problems are explored, including regression, classification, and decision-making problems. The presented works show an increasing trend of new DL techniques that outperform traditional approaches, either through novel architectures or hybrid systems. © 2023 The Institute of Electrical and Electronics Engineers, Inc.