In this thesis, we examine algorithms for the division of labour in a group of robot. The algorithms make no use of direct communication. Instead, they are based only on the interactions among the robots and between the group and the environment. Division of labour is the mechanism that decides how many robots shall be used to perform a task. The efficiency of the group of robots depends in fact on the number of robots involved in a task. If too few robots are used to achieve a task, they might not be successful or might perform poorly. If too many robots are used, it might be a waste of resources. The number of robots to use might be decided a priori by the system designer. More interestingly, the group of robots might autonomously select how many and which robots to use. In this thesis, we study algorithms of the latter type. The robotic literature offers already some solutions, but most of them use a form of direct communication between agents. Direct, or explicit, communication between the robots is usually considered a necessary condition for co-ordination. Recent studies have questioned this assumption. The claim is based on observations of animal colonies, e.g., ants and termites. They can effectively co-operate without directly communicating, but using indirect forms of communication like stigmergy. Because they do not rely on communication, such colonies show robust behaviours at group level, a condition that one wishes also for groups of robots. Algorithms for robot co-ordination without direct communication have been proposed in the last few years. They are interesting not only because they are a stimulating intellectual challenge, but also because they address a situation that might likely occur when using robots for real-world out-door applications. Unfortunately, they are still poorly studied. This thesis helps the understanding and the development of such algorithms. We start from a specific case to learn its characteristics. Then we improve our understandings through comparisons with other solutions, and finally we port everything into another domain. We first study an algorithm for division of labour that was inspired by ants' foraging. We test the algorithm in an application similar to ants' foraging: prey retrieval. We prove that the model used for ants' foraging can be effective also in real conditions. Our analysis allows us to understand the underlying mechanisms of the division of labour and to define some way of measuring it. Using this knowledge, we continue by comparing the ant-inspired algorithm with similar solutions that can be found in the literature and by assessing their differences. In performing these comparisons, we take care of using a formal methodology that allows us to spare resources. Namely, we use concepts of experiment design to reduce the number of experiments with real robots, without losing significance in the results. Finally, we apply and port what we previously learnt into another application: Sensor/Actor Networks (SANETs). We develop an architecture for division of labour that is based on the same mechanisms as the ants' foraging model. Although the individuals in the SANET can communicate, the communication channel might be overloaded. Therefore, the agents of a SANET shall be able to co-ordinate without accessing the communication channel.

Yes / An efficient multi-object evolutionary algorithms are proposed for optimizing frequency characteristics of antennas based on an interfacing created by Matlab environment. This interface makes a link with CST Microwave studio where the electromagnetic investigation of antenna is realized. Very small, compact printed monopole antenna is optimized for ultra- wideband (UWB) applications. Two objective functions are introduced; the first function intends to increase the impedance bandwidth, and second function to tune the antenna to resonate at a particular frequency. The two functions operate in the range of 3.2 to 10.6 GHz and depend on the level of return loss. The computed results provide a set of proper design for UWB system in which the bandwidth achieved is 7.5GHz at the resonance frequency 4.48GHz, including relatively stable gain and radiation patterns across the operating band.

Orientador: Percival Bueno de Araujo / Resumo: Neste trabalho são apresentados quatro métodos de otimização bio-inspirados, Colônia de Abelhas Artificiais, Otimização por Enxame de Partículas, Algoritmo dos Vagalumes e um híbrido aqui denominado por Bee – PSO, que combina particularidades dos outros três. Estes métodos são utilizados no ajuste coordenado dos parâmetros dos controladores Proporcional-Integral e Suplementares de Amortecimento (Estabilizadores de Sistemas de Potência e o conjunto Unified Power Flow Controller – Power Oscillation Damping). O objetivo é inserir amortecimento adicional aos modos oscilatórios de baixa frequência e, consequentemente, garantir a estabilidade do sistema elétrico frente a pequenas perturbações. São considerados três cenários que englobam duas configurações de instalação dos controladores suplementares e duas condições de carregamento, uma fixa e outra variável. Uma formulação por injeções de corrente do dispositivo Unified Power Flow Controller é sugerida e incorporada ao Modelo de Sensibilidade de Corrente, utilizado para representar o sistema elétrico de potência. Análises estática e dinâmica foram realizadas nos sistemas teste Simétrico de Duas Áreas e New England para validar o modelo de injeções de corrente proposto para o Unified Power Flow Controller e determinar qual dos algoritmos apresentados é o mais eficiente no ajuste coordenado dos parâmetros dos controladores. Dos resultados obtidos foi possível concluir que a versão híbrida proposta neste trabalho possui desempenho s... (Resumo completo, clicar acesso eletrônico abaixo) / Abstract: In this work four bio-inspired optimization methods, Artificial Bee Colony, Particle Swarm Optimization, Firefly Algorithm, and a hybrid called Bee – PSO, which combines the characteristics of the other three are presented. These methods are used in the coordinated adjustment of the parameters of Proportional-Integral and Supplementary Damping Controllers (Power System Stabilizers and the Unified Power Flow Controller - Power Oscillation Damping). The goal is to insert additional damping into the low-frequency oscillatory modes and thus ensure the stability of the electrical system against minor disturbances. Three scenarios are considered that include two installation configurations of the supplementary controllers and two charging conditions, one fixed and one variable. A current injection formulation of the Unified Power Flow Controller is suggested and incorporated into the Current Sensitivity Model used to represent the electric power system. Static and dynamic analyzes were performed in the Two-Zone Symmetric and New England test systems to validate the proposed current injection model for the Unified Power Flow Controller and to determine which of the presented algorithms is the most efficient in the coordinated adjustment of the parameters of the controllers. From the results obtained it was possible to conclude that the hybrid version proposed in this work has superior performance in most scenarios analyzed, providing solutions with sufficient damping, even when smal... (Complete abstract click electronic access below) / Doutor

Submitted by LUIS FABIANO BARONE MARTINS null (luis_barone@ig.com.br) on 2017-08-30T14:29:52Z No. of bitstreams: 1 tese-luis-fabiano-barone-martins.pdf: 3400963 bytes, checksum: 6361e93ee21ab8eb5a35cb9dd7d0bd28 (MD5) / Approved for entry into archive by Luiz Galeffi (luizgaleffi@gmail.com) on 2017-08-30T17:52:14Z (GMT) No. of bitstreams: 1 martins_lfb_dr_ilha.pdf: 3400963 bytes, checksum: 6361e93ee21ab8eb5a35cb9dd7d0bd28 (MD5) / Made available in DSpace on 2017-08-30T17:52:14Z (GMT). No. of bitstreams: 1 martins_lfb_dr_ilha.pdf: 3400963 bytes, checksum: 6361e93ee21ab8eb5a35cb9dd7d0bd28 (MD5) Previous issue date: 2017-08-22 / Neste trabalho são apresentados quatro métodos de otimização bio-inspirados, Colônia de Abelhas Artificiais, Otimização por Enxame de Partículas, Algoritmo dos Vagalumes e um híbrido aqui denominado por Bee – PSO, que combina particularidades dos outros três. Estes métodos são utilizados no ajuste coordenado dos parâmetros dos controladores Proporcional-Integral e Suplementares de Amortecimento (Estabilizadores de Sistemas de Potência e o conjunto Unified Power Flow Controller – Power Oscillation Damping). O objetivo é inserir amortecimento adicional aos modos oscilatórios de baixa frequência e, consequentemente, garantir a estabilidade do sistema elétrico frente a pequenas perturbações. São considerados três cenários que englobam duas configurações de instalação dos controladores suplementares e duas condições de carregamento, uma fixa e outra variável. Uma formulação por injeções de corrente do dispositivo Unified Power Flow Controller é sugerida e incorporada ao Modelo de Sensibilidade de Corrente, utilizado para representar o sistema elétrico de potência. Análises estática e dinâmica foram realizadas nos sistemas teste Simétrico de Duas Áreas e New England para validar o modelo de injeções de corrente proposto para o Unified Power Flow Controller e determinar qual dos algoritmos apresentados é o mais eficiente no ajuste coordenado dos parâmetros dos controladores. Dos resultados obtidos foi possível concluir que a versão híbrida proposta neste trabalho possui desempenho superior na maioria dos cenários analisados, fornecendo soluções com amortecimento suficiente, mesmo quando pequenas variações no carregamento do sistema são consideradas. / In this work four bio-inspired optimization methods, Artificial Bee Colony, Particle Swarm Optimization, Firefly Algorithm, and a hybrid called Bee – PSO, which combines the characteristics of the other three are presented. These methods are used in the coordinated adjustment of the parameters of Proportional-Integral and Supplementary Damping Controllers (Power System Stabilizers and the Unified Power Flow Controller - Power Oscillation Damping). The goal is to insert additional damping into the low-frequency oscillatory modes and thus ensure the stability of the electrical system against minor disturbances. Three scenarios are considered that include two installation configurations of the supplementary controllers and two charging conditions, one fixed and one variable. A current injection formulation of the Unified Power Flow Controller is suggested and incorporated into the Current Sensitivity Model used to represent the electric power system. Static and dynamic analyzes were performed in the Two-Zone Symmetric and New England test systems to validate the proposed current injection model for the Unified Power Flow Controller and to determine which of the presented algorithms is the most efficient in the coordinated adjustment of the parameters of the controllers. From the results obtained it was possible to conclude that the hybrid version proposed in this work has superior performance in most scenarios analyzed, providing solutions with sufficient damping, even when small variations in system loading are considered.

Orientador: Akebo Yamakami / Dissertação (mestrado) - Universidade Estadual de Campinas, Faculdade de Engenharia Eletrica e de Computação / Made available in DSpace on 2018-08-11T03:11:24Z (GMT). No. of bitstreams: 1 Rampazzo_PriscilaCristinaBerbert_M.pdf: 1295026 bytes, checksum: ad0738bc161445ec5b9f0db0db565f09 (MD5) Previous issue date: 2008 / Resumo: O objetivo desta dissertação é explorar a utilização de um Sistema Imunológico Artificial, baseado no princípio de Seleção Clonal, na resolução de problemas de Otimização Multiobjetivo. Os Sistemas Imunológicos Artificiais apresentam, em sua estrutura elementar, as principais características requeridas para a resolução de problemas de Otimização Multiobjetivo: exploração, explotação, paralelismo, elitismo, memória, diversidade, mutação e clonagem proporcionais à afinidade e população dinâmica. A abordagem proposta utiliza o conceito de Pareto dominância e factibilidade para identificar os anticorpos (soluções) que devem ser clonados. Nos experimentos, foram consideradas algumas situações importantes que podem aparecer nos problemas reais: presença de restrições (lineares e não-lineares) e formato da Fronteira de Pareto (convexa, côncava, contínua, descontínua, discreta, não-uniforme). Na maioria dos problemas, o algoritmo obteve resultados bons e competitivos quando comparados com as propostas da literatura. Palavras-chave: Otimização Multiobjetivo, Algoritmos Bio-inspirados, Sistemas Imunológicos Artificiais, Seleção Clonal / Abstract: The aim of this work is to explore an Artificial Immune System, based on the Clonal Selection principle, in the solution of Multiobjective Optimization problems. Artificial Immune Systems have, in their elementary structure, the main characteristics required to solve Multiobjective Optimization problems: exploration, exploitation, paralelism, elitism, memory, diversity, mutation and proliferation proportional to the affinity, and dynamic repertorie. The proposed algorithm uses the Pareto dominance concept and feasibility to identify the antibodies (solutions) that must to be cloned. In the experiments, some important situations that occurs in real problems were considered: the presence of constraints (linear and non-linear) and Pareto Front format (convex, concave, continuous, discontinuous, discrete, non-uniforme). In the major part of the problems, the algorithm obtains good and competitive results when compared with approaches from the literature. Keywords: Multiobjective Optimization, Bio-inspired Algorithms, Artificial Immune Systems, Clonal Selection / Mestrado / Telecomunicações e Telemática / Mestre em Engenharia Elétrica

Tesis por compendio / [ES] Las fuentes de energía renovable (ER) permiten una alta disgregación, por lo que hacen posible generar la energía que se utilizará en el mismo sitio de su aprovechamiento. Esto favorece un cambio en la estructura de las redes eléctricas, permitiendo pasar de un esquema de generación centralizado a un esquema distribuido. Sin embargo, las fuentes de ER son altamente dependientes de las condiciones medioambientales como la radiación solar, la nubosidad, el viento, entre otros, por lo que lograr un sistema de generación basado en energías renovables es un reto en la actualidad. Los sistemas de generación que integran fuentes renovables tienen que ser capaces de establecer estrategias de control y gestión de la energía que para hacer un uso eficiente de ella e intentar cubrir la demanda de energía de forma óptima al combinar más de un tipo de fuente y sistema de almacenamiento, siendo posible operar de manera aislada o conectada a la red eléctrica. En la actualidad es de interés el estudio, desarrollo e implementación de sistemas gestores de la energía (SGE) para microrredes eléctricas híbridas, que permitan aumentar su eficiencia, fiabilidad, y disminuir los costes de instalación, operación y mantenimiento. Diversos estudios de investigación han probado múltiples estrategias, desde SGE basados en reglas, algoritmos comparativos, controladores clásicos, y en años recientes, la integración de algoritmos de optimización bio-inspirados e inteligencia artificial. Estos algoritmos han mostrado ser una alternativa interesante a las técnicas clásicas para la solución de problemas de optimización y control en diversos problemas de ingeniería, su aplicación en el ámbito de las microrredes sigue en estudio y en ello se basa este trabajo de investigación. Los algoritmos bio-inspirados se fundamentan en imitar matemáticamente los mecanismos y estrategias que la naturaleza ha implementado a lo largo de millones de años para lograr un equilibrio en su funcionamiento, por ejemplo, imitando el cómo las aves vuelan en parvada buscando alimento, o como las hormigas y los lobos siguen patrones para la búsqueda de su alimento, o como las especies llevan a cabo mecanismos de cruce con el objetivo de mejorar su raza haciéndolas una especie optimizada y mejorando su supervivencia. Por tanto, se puede hacer una analogía con los sistemas artificiales para la mejora de controladores y diseño de sistemas en microrredes eléctricas. En este trabajo de investigación se muestra el modelo y desarrollo de un sistema de gestión óptima para una microrred empleando algoritmos bio-inspirados con el objetivo de mejorar su desempeño, partiendo desde el control primario, con la mejora de los convertidores de potencia, hasta el control terciario con las transacciones energéticas de la microrred. Se exploran diversos algoritmos, evaluando su desempeño. Los resultados para las diferentes etapas de esta investigación se encuentran plasmados en cuatro diferentes publicaciones científicas que se detallan en el Capítulo 2 del presente documento, donde se explica la metodología y los principales resultados y hallazgos para cada una de ellas. / [CA] Les fonts d'energia renovables (ER) permeten una alta desagregació, pel que fan possible generar l'energia que s'utilitzarà en el mateix lloc del seu aprofitament. Això afavoreix un canvi en l'estructura de les xarxes elèctriques, permetent passar d'un esquema de generació centralitzat a un esquema distribuït. No obstant, les fonts d'ER són altament dependents de les condicions mediambientals com la radiació solar, la nuvolositat, el vent, entre altres; pel que aconseguir un sistema de generació basat en energies renovables és un repte. Els sistemes de generació que integren energies renovables han de ser capaços de: establir estratègies de control i gestió de l'energia que es genera per fer un ús eficient d'ella i intentar cobrir la demanda d'energia de la millor manera possible al combinar més d'un tipus de font d'energia, i sistemes d'emmagatzemament. Aquest esquema es coneix com a microxarxa elèctrica, la qual és capaç d'operar de manera aïllada de la xarxa elèctrica principal, o de manera interconnectada. Actualment s'està interessant en l'estudi, desenvolupament i implementació de sistemes gestors de l'energia (SGE) per a microxarxes elèctriques híbrides, que permeten augmentar la seua eficiència, fiabilitat i reduir els costos de la seua instal·lació i d'operació i manteniment. S'han provat múltiples estratègies, des de SGE basats en regles, algorismes comparatius, controladors clàssics i, en anys recents, la integració d'algorismes d'optimització bio-inspirats i intel·ligència artificial. Aquests algorismes han demostrat ser una alternativa interessant a les tècniques clàssiques per a la solució de problemes d'optimització i control en diversos problemes d'enginyeria, la seua aplicació en l'àmbit de les microxarxes continua en estudi. Els algorismes bio-inspirats es basen en imitar matemàticament els mecanismes i estratègies que la Natura ha implementat al llarg de milions d'anys per aconseguir equilibri en el seu funcionament, per exemple, imitant com les aus volen en ramat buscant menjar, o com les formigues i els llops segueixen patrons per a la recerca del seu menjar, o com les espècies porten a terme mecanismes de creuament amb mira a millorar la seua raça fent-les una espècie més apta per a la supervivència;, el qual es pot fer una analogia a sistemes artificials per a la millora de controladors i disseny de sistemes en microxarxes elèctriques. En aquest treball de recerca es mostra el model i desenvolupament d'un sistema de gestió òptima per a una microxarxa emprant algorismes bio-inspirats amb l'objectiu de millorar el seu rendiment, partint des del control primari, amb la millora dels convertidors de potència, fins al control terciari amb les transaccions energètiques de la microxarxa. S'exploren diversos algorismes, avaluant el seu rendiment. Els resultats per a les diferents etapes d'aquesta recerca es troben plasmats en quatre diferents publicacions científiques que es detallen al Capítol 2 del present document, on s'explica la metodologia i els principals resultats i troballes per a cada una d'elles. / [EN] Renewable energy sources (RES) allow for high disaggregation, making it possible to generate energy at the site of its use. This favors a change in the structure of electrical grids, allowing for a transition from a centralized generation scheme to a distributed scheme. However, RES are highly dependent on environmental conditions such as solar radiation, cloudiness, wind, among others, making the creation of a renewable energy generation system a challenge. Generation systems that integrate renewable energies must be able to establish control and energy management strategies to make efficient use of the energy generated and try to meet the energy demand in the best possible way by combining more than one type of energy source and storage systems. This scheme is known as a microgrid, which is capable of operating independently from the main electrical grid or interconnecting with it. Currently, the study, development, and implementation of energy management systems (EMS) for hybrid microgrids are of interest in order to increase their efficiency, reliability, and reduce installation, operation, and maintenance costs. Multiple strategies have been tested, including rule-based EMS, comparative algorithms, classical controllers, and in recent years, the integration of bio-inspired optimization algorithms and artificial intelligence. These algorithms have shown to be an interesting alternative to classical techniques for solving optimization and control problems in various engineering problems, although their application in the field of microgrids is still under study. Bio-inspired algorithms are based on mathematically imitating the mechanisms and strategies that Nature has implemented over millions of years to achieve balance in its operation, for example, by imitating how birds fly in flocks in search of food, or how ants and wolves follow patterns to search for food, or how species carry out crossing mechanisms in order to improve their breed and make them more suitable for survival; in other words, they are based on how Nature optimizes its resources to prosper. Therefore, an analogy can be made with artificial systems for improving controllers and designing systems in microgrids. In this research work, the model and development of an optimal management system for a microgrid using bio-inspired algorithms is presented with the aim of improving its performance, starting from primary control, with the improvement of power converters, to tertiary control with the energy transactions of the microgrid. Various algorithms are explored, evaluating their performance. The results for the different stages of this research are reflected in four different scientific publications that are detailed in Chapter 2 of this document, where the methodology and main results and findings for each of them are explained. / The authors wish to acknowledge the National Council of Science and Technology of Mexico (CONACYT) for funding this work through the Ph.D. scholarship number 486670. The authors would also thank the Institute of Energy Engineering of the Polytechnic University of Valencia, Spain, and the Department of Water and Energy Studies of the University of Guadalajara, Mexico, for all their support and collaboration. This study has also been supported by Food and Agriculture Organization of the United Nations through the project “Design of a Hybrid Renewable Microgrid System”. / Águila León, J. (2023). Modelo y desarrollo de un sistema de gestión óptima para una microrred empleando algoritmos bio-inspirados [Tesis doctoral]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/196747 / Compendio

Une tâche importante qui a été étudiée dans le contexte de données à forte dimensionnalité est la tâche connue sous le nom de subspace clustering. Le subspace clustering est généralement reconnu comme étant plus compliqué que le clustering standard, étant donné que cette tâche vise à détecter des groupes d’objets similaires entre eux (clusters), et qu’en même temps elle vise à trouver les sous-espaces où apparaissent ces similitudes. Le subspace clustering, ainsi que le clustering traditionnel ont été récemment étendus au traitement de flux de données en mettant à jour les modèles de clustering de façon incrémentale. Les différents algorithmes qui ont été proposés dans la littérature, reposent sur des bases algorithmiques très différentes. Parmi ces approches, les algorithmes évolutifs ont été sous-explorés, même si ces techniques se sont avérées très utiles pour traiter d’autres problèmes NP-difficiles. L’objectif de cette thèse a été de tirer parti des nouvelles connaissances issues de l’évolution afin de concevoir des algorithmes évolutifs qui traitent le problème du subspace clustering sur des jeux de données statiques ainsi que sur des flux de données dynamiques. Chameleoclust, le premier algorithme développé au cours de ce projet, tire partie du grand degré de liberté fourni par des éléments bio-inspirés tels qu’un génome de longueur variable, l’existence d’éléments fonctionnels et non fonctionnels et des opérateurs de mutation incluant des réarrangements chromosomiques. KymeroClust, le deuxième algorithme conçu dans cette thèse, est un algorithme de k-medianes qui repose sur un mécanisme évolutif important: la duplication et la divergence des gènes. SubMorphoStream, le dernier algorithme développé ici, aborde le problème du subspace clustering sur des flux de données dynamiques. Cet algorithme repose sur deux mécanismes qui jouent un rôle clef dans l’adaptation rapide des bactéries à des environnements changeants: l’amplification de gènes et l’absorption de matériel génétique externe. Ces algorithmes ont été comparés aux principales techniques de l’état de l’art, et ont obtenu des résultats compétitifs. En outre, deux applications appelées EvoWave et EvoMove ont été développés pour évaluer la capacité de ces algorithmes à résoudre des problèmes réels. EvoWave est une application d’analyse de signaux Wi-Fi pour détecter des contextes différents. EvoMove est un compagnon musical artificiel qui produit des sons basés sur le clustering des mouvements d’un danseur, décrits par des données provenant de capteurs de déplacements. / An important task that has been investigated in the context of high dimensional data is subspace clustering. This data mining task is recognized as more general and complicated than standard clustering, since it aims to detect groups of similar objects called clusters, and at the same time to find the subspaces where these similarities appear. Furthermore, subspace clustering approaches as well as traditional clustering ones have recently been extended to deal with data streams by updating clustering models in an incremental way. The different algorithms that have been proposed in the literature, rely on very different algorithmic foundations. Among these approaches, evolutionary algorithms have been under-explored, even if these techniques have proven to be valuable addressing other NP-hard problems. The aim of this thesis was to take advantage of new knowledge from evolutionary biology in order to conceive evolutionary subspace clustering algorithms for static datasets and dynamic data streams. Chameleoclust, the first algorithm developed in this work, takes advantage of the large degree of freedom provided by bio-like features such as a variable genome length, the existence of functional and non-functional elements and mutation operators including chromosomal rearrangements. KymeroClust, our second algorithm, is a k-medians based approach that relies on the duplication and the divergence of genes, a cornerstone evolutionary mechanism. SubMorphoStream, the last one, tackles the subspace clustering task over dynamic data streams. It relies on two important mechanisms that favor fast adaptation of bacteria to changing environments, namely gene amplification and foreign genetic material uptake. All these algorithms were compared to the main state-of-the-art techniques, obtaining competitive results. Results suggest that these algorithms are useful complementary tools in the analyst toolbox. In addition, two applications called EvoWave and EvoMove have been developed to assess the capacity of these algorithms to address real world problems. EvoWave is an application that handles the analysis of Wi-Fi signals to detect different contexts. EvoMove, the second one, is a musical companion that produces sounds based on the clustering of dancer moves captured using motion sensors.

In this thesis, we examine algorithms for the division of labour in a group of robot. The algorithms make no use of direct communication. Instead, they are based only on the interactions among the robots and between the group and the environment.<p><p>Division of labour is the mechanism that decides how many robots shall be used to perform a task. The efficiency of the group of robots depends in fact on the number of robots involved in a task. If too few robots are used to achieve a task, they might not be successful or might perform poorly. If too many robots are used, it might be a waste of resources. The number of robots to use might be decided a priori by the system designer. More interestingly, the group of robots might autonomously select how many and which robots to use. In this thesis, we study algorithms of the latter type.<p><p>The robotic literature offers already some solutions, but most of them use a form of direct communication between agents. Direct, or explicit, communication between the robots is usually considered a necessary condition for co-ordination. Recent studies have questioned this assumption. The claim is based on observations of animal colonies, e.g. ants and termites. They can effectively co-operate without directly communicating, but using indirect forms of communication like stigmergy. Because they do not rely on communication, such colonies show robust behaviours at group level, a condition that one wishes also for groups of robots. Algorithms for robot co-ordination without direct communication have been proposed in the last few years. They are interesting not only because they are a stimulating intellectual challenge, but also because they address a situation that might likely occur when using robots for real-world out-door applications. Unfortunately, they are still poorly studied.<p><p>This thesis helps the understanding and the development of such algorithms. We start from a specific case to learn its characteristics. Then we improve our understandings through comparisons with other solutions, and finally we port everything into another domain.<p><p>We first study an algorithm for division of labour that was inspired by ants' foraging. We test the algorithm in an application similar to ants' foraging: prey retrieval. We prove that the model used for ants' foraging can be effective also in real conditions. Our analysis allows us to understand the underlying mechanisms of the division of labour and to define some way of measuring it.<p><p>Using this knowledge, we continue by comparing the ant-inspired algorithm with similar solutions that can be found in the literature and by assessing their differences. In performing these comparisons, we take care of using a formal methodology that allows us to spare resources. Namely, we use concepts of experiment design to reduce the number of experiments with real robots, without losing significance in the results.<p><p>Finally, we apply and port what we previously learnt into another application: Sensor/Actor Networks (SANETs). We develop an architecture for division of labour that is based on the same mechanisms as the ants' foraging model. Although the individuals in the SANET can communicate, the communication channel might be overloaded. Therefore, the agents of a SANET shall be able to co-ordinate without accessing the communication channel. / Doctorat en sciences appliquées / info:eu-repo/semantics/nonPublished

[pt] Problemas reais de engenharia mecânica podem compreender tarefas de i) otimização multi-objetivo (MO) ou ii) regressão, classificação e predição. Os métodos baseados em inteligência artificial (AI) são bastante difundidos na resolução desses problemas por i) demandarem menor custo computacional e informações do domínio do problema para a resolução de uma MO, quando comparados com métodos de programação matemática, por exemplo; e ii) apresentarem melhores resultados com estrutura mais simples, adaptabilidade e interpretabilidade, em contraste com outros métodos. Sendo assim, o presente trabalho busca i) otimizar um controle proporcional-integral-derivativo (PID) aplicado a um sistema de frenagem anti-travamento de rodas (ABS) e o projeto de trocadores de calor de placas aletadas (PFHE) e casco-tubo (STHE) através de métodos de otimização baseados AI, buscando o desenvolvimento de novas versões dos métodos aplicados, e.g. multi-objective salp swarm algorithm (MSSA) e multi-objective heuristic Kalman algorithm (MOHKA), que melhorem a performance da otimização; ii) desenvolver um sistema de detecção de vazamento em dutos (LDS) sensível ao roubo de combustível a partir do treinamento de árvores de decisão (DTs) com features baseadas no tempo e na análise de componentes principais (PCA), ambas exraídas de dados de transiente de pressão de operação normal do duto e de roubo de combustível; iii) constituir um guia de aplicação para problemas de MO de controle e projeto, processo de extração de features e treinamento de classificadores baseados em aprendizado de máquina (MLCs), através de aprendizado supervisionado; e, por fim iv) demonstrar o potencial das técnicas baseadas em AI. / [en] Real-world mechanical engineering problems may comprise tasks of i) multi-objective optimization (MO) or ii) regression, classification and prediction. The use of artificial intelligence (AI) based methods for solving these problems are widespread for i) demanding less computational cost and problem domain information to solve the MO, when compared with mathematical programming for an example; and ii) presenting better results with simpler structure, adaptability and interpretability, in contrast to other methods. Therefore, the present work seeks to i) optimize a proportional-integral-derivative control (PID) applied to an anti-lock braking system (ABS) and the heat exchanger design of plate-fin (PFHE) and shell-tube (STHE) types through AI based optimization methods, seeking to develop new versions of the applied methods, e.g. multi-objective salp swarm algorithm (MSSA) and multi-objective heuristic Kalman algorithm (MOHKA), which enhance the optimization performance; ii) develop a pipeline leak detection system (LDS) sensitive to fuel theft by training decision trees (DTs) with features based on time and principal component analysis (PCA), both extracted from pressure transient data of regular pipeline operation and fuel theft; iii) constitute an application guide for control and design MO problems, feature extraction process and machine learning classifiers (MLCs) training through supervised learning; and, finally, iv) demonstrate the potential of AI-based techniques.

Tesis por compendio / [ES] La energía, la comunicación y la informática son componentes fundamentales de la sociedad moderna, ya que sientan las bases para el desarrollo tecnológico y el crecimiento económico. La estrecha interrelación entre estos pilares se ha hecho cada vez más evidente en los últimos años, a medida que los avances en computación y análisis de datos han permitido nuevos enfoques de gestión y sostenibilidad de la energía. En este contexto, el uso eficiente de la energía se ha convertido en un objetivo clave para los investigadores, los responsables políticos y las empresas por igual. Al aprovechar el poder de las técnicas informáticas y de aprendizaje automático (ML), es posible destacar los desafíos de asegurar los sistemas de energía y optimizar el uso de la energía, lo que lleva a la necesidad de técnicas avanzadas como algoritmos bio-inspirados y redes neuronales. Esta tesis doctoral tiene como objetivo analizar los programas y estrategias de gestión de la carga, el consumo y la demanda en el panorama energético actual. El núcleo central presenta un estudio exhaustivo sobre la integración de algoritmos bio-inspirados, como la optimización de enjambres de partículas (PSO) y los modelos de redes neuronales artificiales (ANN) en los sistemas de gestión de la carga para hacer frente a los retos de la gestión de la carga y utilizar la energía de forma eficiente y segura. El cuerpo principal de esta tesis comprende tres publicaciones científicas, cada una de las cuales corresponde a una etapa distinta dentro del marco general de investigación de este estudio: la primera etapa propone un sistema de monitorización de bajo coste para aplicaciones energéticas que introduce un sistema SCADA basado en web rentable que era un del 80% más barato que una solución similar. La arquitectura de bajo coste propuesta, diseñada para bancos de pruebas de microrredes, ofrece monitorización en tiempo real, accesibilidad remota y control fácil de usar para aplicaciones académicas y de investigación. La segunda etapa combina la optimización híbrida de enjambre de partículas (PSO) en cascada con redes neuronales feed-forward para pronosticar y optimizar con precisión la demanda de energía en una microrred en AC, mejorando la integración de fuentes de energía renovables como gasificación de biomasa. Los resultados muestran que el modelo PSO-ANN propuesto tiene un rendimiento un 23,2% mejor en términos de MSE que los modelos de RNA de retropropagación feed-forward (FF-BP) y propagación directa en cascada (CF-P). La tercera y última etapa se centró en un sistema inteligente de gestión de la carga reforzado con criptografía híbrida para garantizar la comunicación protegida y la privacidad de los datos, abordando así de manera efectiva los desafíos de seguridad energética en entornos residenciales. Los resultados mostraron que el modelo propuesto de Gestión de Carga aplicado a Sistemas Residenciales de Seguridad (SRS-LM) fue un 37% mejor en rendimiento (costo de energía, utilización de energía, tiempo computacional) y con una reducción de carga máxima del 60% en comparación con un modelo de Medidor de Energía Inteligente Universal (USEM). / [CA] L'energia, la comunicació i la informàtica són components fonamentals de la societat moderna, ja que establixen les bases per al desenvolupament tecnològic i el creixement econòmic. L'estreta interrelació entre estos pilars s'ha fet cada vegada més evident en els últims anys, a mesura que els avanços en computació i anàlisi de dades han permés nous enfocaments de gestió i sostenibilitat de l'energia. En este context, l'ús eficient de l'energia s'ha convertit en un objectiu clau per als investigadors, els responsables polítics i les empreses per igual. En aprofitar el poder de les tècniques informàtiques i d'aprenentatge automàtic (ML), és possible destacar els desafiaments d'assegurar els sistemes d'energia i optimitzar l'ús de l'energia, la qual cosa porta a la necessitat de tècniques avançades com a algorismes bio-inspirats i xarxes neuronals. Esta tesi doctoral té com a objectiu analitzar els programes i estratègies de gestió de la càrrega, el consum i la demanda en el panorama energètic actual. El nucli central presenta un estudi exhaustiu sobre la integració d'algorismes bio-inspirats, com l'optimització d'eixams de partícules (PSO) i els models de xarxes neuronals artificials (ANN) en els sistemes de gestió de la càrrega per a fer front als reptes de la gestió de la càrrega i utilitzar l'energia de manera eficient i segura. El cos principal d'esta tesi comprén tres publicacions científiques, cadascuna de les quals correspon a una etapa diferent dins del marc general d'investigació d'este estudi: la primera etapa proposa un sistema de monitoratge de baix cost per a aplicacions energètiques que introduïx un sistema SCADA basat en web rendible que era un del 80% més barat que una solució similar. L'arquitectura de baix cost proposada, dissenyada per a bancs de proves de microxarxes, oferix monitoratge en temps real, accessibilitat remota i control fàcil d'usar per a aplicacions acadèmiques i d'investigació. La segona etapa combina l'optimització híbrida d'eixam de partícules (PSO) en cascada amb xarxes neuronals feed-forward per a pronosticar i optimitzar amb precisió la demanda d'energia en una microxarxa en AC, millorant la integració de fonts d'energia renovables com a gasificació de biomassa. Els resultats mostren que el model PSO-ANN proposat té un rendiment un 23,2% millor en termes de MSE que els models d'RNA de retropropagació feed-forward (FF-BP) i propagació directa en cascada (CF-P). La tercera i última etapa es va centrar en un sistema intel·ligent de gestió de la càrrega reforçat amb criptografia híbrida per a garantir la comunicació protegida i la privacitat de les dades, abordant així de manera efectiva els desafiaments de seguretat energètica en entorns residencials. Els resultats van mostrar que el model proposat de Gestió de Càrrega aplicat a Sistemes Residencials de Seguretat (SRS-LM) va ser un 37% millor en rendiment (cost d'energia, utilització d'energia, temps computacional) i amb una reducció de càrrega màxima del 60% en comparació amb un model de Mesurador d'Energia Intel·ligent Universal (USEM). / [EN] Energy, communication, and computing are critical components of modern society, providing the foundation for technological development and economic growth. The close interrelation between these pillars has become increasingly apparent in recent years, as computing and data analysis advances have enabled new energy management and sustainability approaches. In this context, efficient energy usage has become a key focus for researchers, policymakers, and businesses alike. By harnessing the power of computing and machine learning (ML) techniques, it is possible to highlight the challenges of securing energy systems and optimizing energy usage, leading to the need for advanced techniques such as bio-inspired algorithms and neural networks. This doctoral thesis aims to analyse load consumption and demand management programs and strategies in the current energy landscape. The central core presents an study on integrating bio-inspired algorithms, such as particle swarm optimization (PSO) and artificial neural networks (ANN) models in load management systems to meet load management challenges and use energy efficiently and securely. The main body of this thesis comprises three scientific publications, each corresponding to a distinct stage within the overarching research framework of this study: the first stage covers the proposal of a low-cost architecture in energy systems introducing a cost-effective web-based SCADA system that was over 80% cheaper than a similar solution. The proposed low-cost architecture, tailored for microgrid testbeds, offers real-time monitoring, remote accessibility, and user-friendly control for academic and research applications. The second stage combined a cascade hybrid Particle Swarm Optimization (PSO) with feed-forward neural networks to accurately forecast and optimize energy demand in an AC microgrid, notably enhancing the integration of renewable energy sources like biomass gasification. The results showed that the proposed PSO-ANN model performs 23.2% better in terms of MSE than Feedforward Backpropagation (FF-BP) and Cascade forward propagation (CF-P) ANN models. The third and final stage focused on a smart load management system fortified with hybrid cryptography to ensure protected communication and data privacy, thereby effectively addressing energy security challenges in residential settings. Results showed that the proposed Security Residential System Load Management (SRS-LM) model was 37% better in performance (power cost, power utilization, computational time) and with a 60% peak load reduction compared to a Universal Smart Energy Meter (USEM) model. / Chiñas Palacios, CD. (2024). Bio-Inspired Algorithms and Artificial Neural Networks Applied to Smart Load Management Systems to Optimize Energy Usage [Tesis doctoral]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/203128 / Compendio