Download Free Environment Driven Distributed Evolutionary Adaptation For Collective Robotic Systems Book in PDF and EPUB Free Download. You can read online Environment Driven Distributed Evolutionary Adaptation For Collective Robotic Systems and write the review.

Cette thèse décrit une partie du travail effectué dans le cadre du projet européen Symbrion 1 . Ce projet vise à la réalisation de tâches complexes nécessitant la coopération de multiples robots dans un cadre de robotique en essaim (au moins 100 robots opérant ensemble). De multiples problèmes sont étudiés par le projet dont : l'auto-assemblage de robots en structures complexes et l'auto-organisation d'un grand nombre de robots afin de réaliser une tâche commune. Le principal sujet porte sur les mécanismes d'auto-adaptation pour la robotique modulaire et en essaim, avec un intérêt pour des capacités de forte coordination et de coopération à l'échelle de l'essaim.Les difficultés rencontrées dans la réalisation de ce projet sont dues à l'utilisation de robots dans des environnements ouverts restant inconnus jusqu'à la phase de déploiement. Puisque les conditions d'opérations ne peuvent être prédites à l'avance, des algorithmes d'apprentissage en ligne doivent être utilisés pour élaborer les comportements utilisés. Lorsqu'un grand nombre de robots sont utilisés, plusieurs considérations doivent être prise en compte : capacité de communication réduite, faible mémoire, faible capacité de calcul. Par conséquent les algorithmes d'apprentissage en ligne doivent être distribués à travers l'essaim.De multiples approches ont déjà été proposées pour faire face aux problèmes posés par l'apprentissage en ligne décentralisé de comportements robotiques, parmi lesquels la robotique probabiliste, l'apprentissage par renforcement, et la robotique évolutionnaire. Cependant, le problème abordé dans le cadre de cette thèse se caractérise par le fait que l'on considère un groupe de robots (en lieu et place d'un seul et unique robot). De plus, dû à la nature ouverte de l'environnement, il n'est pas possible de supposer que l'ingénieur humain ait les connaissances nécessaires pour définir les éléments indispensables aux processus d'apprentissage.Assurer l'intégrité de l'essaim est placé en tant que premier élément d'une feuille de route visant à définir un ensemble d'étapes nécessaires à la réalisation d'une tâche par un groupe de robot dans un environnement ouvert :- Étape 1 : Assurer l'intégrité de l'essaim.- Étape 2 : Maintenir les robots disponibles en tant que service à l'utilisateur.- Étape 3 : Réaliser la tâche définie par l'utilisateur.Dans le cadre de cette thèse nous travaillons à la réalisation de l'étape 1 de cette feuille de route, et assumons l'hypothèse de travail suivante :Hypothèse de travail : Dans un cadre de robotique collective en environnement ouvert, la réalisation d'une tâche définie par l'utilisateur implique tout d'abord un comportement auto-adaptatif.Le sujet de cette thèse est la réalisation de solutions algorithmiques décentralisées pouvant garantir l'in- tégrité d'un essaim de robots en environnement ouvert lorsque un système robotique collectif utilise une communication locale. La principale difficulté à sa résolution est le besoin de prendre en compte l'envi- ronnement. En effet, en fonction de l'environnement courant, les robots peuvent avoir à démontrer une grande variété de comportements à l'échelle globale comme la coopération, la spécialisation, l'altruisme, ou la division du travail.Dans cette thèse nous introduisons et définissons le problème de l'Adaptation Evolutionnaire Distribuée Guidée par l'Environnement. Nous proposons un algorithme pour résoudre ce problem. Cet algorithme a été validé aussi bien en simulation que sur des robots réels. Il a été utilisé pour étudier le problème de l'auto-adaptation dans les environnements suivants :- Environnement où l'émergence de consensus comportementaux est nécessaire.- Environnements où la robustesse face à des changements environnementaux est nécessaires.- Environnements où des comportements altruistes sont nécessaires.
Robot swarms are systems composed of a large number of rather simple robots. Due to the large number of units, these systems, have good properties concerning robustness and scalability, among others. However, it remains generally difficult to design controllers for such robotic systems, particularly due to the complexity of inter-robot interactions. Consequently, automatic approaches to synthesize behavior in robot swarms are a compelling alternative. In this thesis, we focus on online behavior adaptation in a swarm of robots using distributed Embodied Evolutionary Robotics (EER) methods. To this end, we provide three main contributions: (1) We investigate the influence of task-driven selection pressure in a swarm of robotic agents using a distributed EER approach. We evaluate the impact of a range of selection pressure strength on the performance of a distributed EER algorithm. The results show that the stronger the task-driven selection pressure, the better the performances obtained when addressing given tasks. (2) We investigate the evolution of collaborative behaviors in a swarm of robotic agents using a distributed EER approach. We perform a set of experiments for a swarm of robots to adapt to a collaborative item collection task that cannot be solved by a single robot. Our results show that the swarm learns to collaborate to solve the task using a distributed approach, and we identify some inefficiencies regarding learning to choose actions. (3) We propose and experimentally validate a completely distributed mechanism that allows to learn the structure and parameters of the robot neurocontrollers in a swarm using a distributed EER approach, which allows for the robot controllers to augment their expressivity. Our experiments show that our fully-decentralized mechanism leads to similar results as a mechanism that depends on global information.
This book constitutes the refereed proceedings of the International Conference on the Applications of Evolutionary Computation, EvoApplications 2012, held in Málaga, Spain, in April 2012, colocated with the Evo* 2012 events EuroGP, EvoCOP, EvoBIO, and EvoMUSART. The 54 revised full papers presented were carefully reviewed and selected from 90 submissions. EvoApplications 2012 consisted of the following 11 tracks: EvoCOMNET (nature-inspired techniques for telecommunication networks and other parrallel and distributed systems), EvoCOMPLEX (algorithms and complex systems), EvoFIN (evolutionary and natural computation in finance and economics), EvoGAMES (bio-inspired algorithms in games), EvoHOT (bio-inspired heuristics for design automation), EvoIASP (evolutionary computation in image analysis and signal processing), EvoNUM (bio-inspired algorithms for continuous parameter optimization), EvoPAR (parallel implementation of evolutionary algorithms), EvoRISK (computational intelligence for risk management, security and defense applications), EvoSTIM (nature-inspired techniques in scheduling, planning, and timetabling), and EvoSTOC (evolutionary algorithms in stochastic and dynamic environments).
The book is a valuable reference work for students, researchers, academics, and industry practitioners interested in the latest scientific and technological advances across the conference topics. The CSA 2022 proceedings provide a collection of new ideas, original research findings, and experimental results in the field of computer science covering: artificial intelligence, data science, computer networks and security, information systems, software engineering, and computer graphics.
The two volumes LNCS 9107 and 9108 constitute the proceedings of the International Work-Conference on the Interplay Between Natural and Artificial Computation, IWINAC 2015, held in Elche, Spain, in June 2015. The total of 103 contributions was carefully reviewed and selected from 190 submissions during two rounds of reviewing and improvement. The papers are organized in two volumes, one on artificial computation and biology and medicine, addressing topics such as computational neuroscience, neural coding and neuro-informatics, as well as computational foundations and approaches to the study of cognition. The second volume deals with bioinspired computation in artificial systems; topics alluded are bio-inspired circuits and mechanisms, bioinspired programming strategies, and bioinspired engineering AI&KE.
This book focuses on the emergence of creative ideas from cognitive and social dynamics. In particular, it presents data, models, and analytical methods grounded in a network dynamics approach. It has long been hypothesized that innovation arises from a recombination of older ideas and concepts, but this has been studied primarily at an abstract level. In this book, we consider the networks underlying innovation – from the brain networks supporting semantic cognition to human networks such as brainstorming groups or individuals interacting through social networks – and relate the emergence of ideas to the structure and dynamics of these networks. Methods described include experimental studies with human participants, mathematical evaluation of novelty from group brainstorming experiments, neurodynamical modeling of conceptual combination, and multi-agent modeling of collective creativity. The main distinctive features of this book are the breadth of perspectives considered, the integration of experiments with theory, and a focus on the combinatorial emergence of ideas.
This book presents recent advances in automated machine learning (AutoML) and automated algorithm design and indicates the future directions in this fast-developing area. Methods have been developed to automate the design of neural networks, heuristics and metaheuristics using techniques such as metaheuristics, statistical techniques, machine learning and hyper-heuristics. The book first defines the field of automated design, distinguishing it from the similar but different topics of automated algorithm configuration and automated algorithm selection. The chapters report on the current state of the art by experts in the field and include reviews of AutoML and automated design of search, theoretical analyses of automated algorithm design, automated design of control software for robot swarms, and overfitting as a benchmark and design tool. Also covered are automated generation of constructive and perturbative low-level heuristics, selection hyper-heuristics for automated design, automated design of deep-learning approaches using hyper-heuristics, genetic programming hyper-heuristics with transfer knowledge and automated design of classification algorithms. The book concludes by examining future research directions of this rapidly evolving field. The information presented here will especially interest researchers and practitioners in the fields of artificial intelligence, computational intelligence, evolutionary computation and optimisation.
Self-contained and unified in presentation, this invaluable book provides a broad introduction to the fascinating subject of many-body collective systems with adapting and evolving agents. The coverage includes game theoretic systems, multi-agent systems, and large-scale socio-economic systems of individual optimizing agents. The diversity and scope of such systems have been steadily growing in computer science, economics, social sciences, physics, and biology.
This book constitutes the refereed proceedings of the 17th Portuguese Conference on Artificial Intelligence, EPIA 2015, held in Coimbra, Portugal, in September 2015. The 45 revised full papers presented together with 36 revised short papers were carefully reviewed and selected from a total of 131 submissions. EPIA 2015, following the standard EPIA format, covers a wide range of AI topics as follows: ambient intelligence and affective environments, artificial Intelligence in medicine, artificial intelligence in transportation systems, artificial life and evolutionary algorithms, computational methods in bioinformatics and systems biology, general artificial intelligence, intelligent information systems, intelligent robotics, knowledge discovery and business intelligence, multi-agent systems: theory and applications, social simulation and modelling, text mining and applications.