La maîtrise de la qualité est aujourd’hui indispensable pour satisfaire les exigences des clients. Bien qu’il existe différentes méthodes et outils spécialement développés pour manager la qualité des systèmes ou des processus au sein des entreprises, il est difficile d’établir le lien entre la maîtrise des nombreux processus d’une entreprise et la qualité produit pour un système en service. Dans ce contexte, une thèse CIFRE a été menée au sein de MBDA, entreprise qui conçoit, développe et produit des systèmes d’armes. La problématique de cette thèse est de développer une méthodologie permettant de produire une définition de la qualité produit d’un système complexe qui soit valable tout au long de son cycle de vie, et permettant la construction de modèles de prévision de la qualité produit en utilisation lors du développement et de la production. Notre contribution consiste en une méthodologie en deux phases. La première phase permet d’établir une définition de la qualité produit des systèmes complexes du point de vue du client adaptée au contexte industriel en définissant plusieurs facteurs qualité produit qui soit valable pour toutes les phases du cycle de vie des systèmes. La deuxième phase permet de construire des modèles de prévision de la qualité produit qui permettent d’obtenir une évaluation de la qualité tout au long du cycle de vie des systèmes et d’établir une prévision de ce que sera la qualité en utilisation. Les deux phases de la méthodologie reposent sur l’exploitation d’avis d’experts afin de permettre son utilisation sans disposer d’une quantité importante de données. Les modèles construits ont été testés pour des systèmes développés par MBDA. / Today, quality control is essential to satisfy customer requirements. Although there are different methods and tools specially developed to manage the quality of systems or processes within companies, it is difficult to establish the link between management of a company's many processes and product quality for a system in service. In this context, a CIFRE thesis was conducted within MBDA, a company that designs, develops and produces weapons systems. The problem of this thesis is to develop a methodology allowing to produce a definition of the product quality of a complex system which is valid throughout its life cycle, and allowing the construction of models to predict the product quality in use during development and production. Our contribution consists of a two-phase methodology. The first phase makes it possible to establish a definition of the product quality of complex systems from the customer's point of view adapted to the industrial context by defining several product quality factors that are relevant for all phases of the systems life cycle. The second phase builds product quality prediction models that provide a life-cycle quality assessment of the systems and a forecast of what the quality will be in use. Both phases of the methodology rely on the use of experts' judgement to enable its use without a significant amount of data. The models built have been tested for systems developed by MBDA.

A histórica desconsideração da variável ambiental dentro da concepção teórica e dos modelos de crescimento econômico revela um posicionamento ideológico muito claro: a resistência na aceitação do ambiente como fator limitante ou mesmo como elemento estratégico do ponto de vista da riqueza de uma nação ou região. Para além das questões mais frequentemente debatidas, que associam os elementos do ambiente a meros insumos necessários aos processos produtivos, a consideração de conceitos não usuais no arcabouço teórico da economia tradicional, como serviços ecossistêmicos, resiliência, entropia e histerese, teria condições de ampliar e relativizar a interpretação de uma série de premissas e dogmas da ciência econômica tradicional. O conceito de capital natural, neste sentido, apresenta-se como fundamental e estratégico, uma vez que permite acomodar toda a complexidade inerente à dimensão ambiental e relacioná-la com o sistema socioeconômico, adequando e balizando o enquadramento da questão da sustentabilidade. Ainda, o presente trabalho parte da definição fundamentada de que ambos os sistemas - econômico e ambiental - são essencialmente complexos e, que os efeitos das relações entre os mesmos não são triviais e possuem altos níveis de incerteza associados à sua dinâmica. Dentro dessa orientação, o trabalho se propôs a realizar uma reflexão sobre a sustentabilidade sob a ótica dos sistemas complexos, por meio de uma revisão bibliográfica crítica e de um exercício de modelagem baseada em agentes para a simulação do crescimento econômico considerando a variável ambiental. As análises realizadas indicam que a incorporação de novos conceitos oriundos dos sistemas complexos poderiam estabelecer um novo suporte para a análise de políticas macroeconômicas de crescimento, da sustentabilidade e, em última instância, contribuir com o fortalecimento de premissas básicas da economia ecológica. / The historical disregard of the environmental issue in the theoretical conception and the economic growth models reveals a clear ideological positioning: the resistance to accept the environment as a limiting factor or as a strategic element from the point of view of nations or regions wealth. Beyond the frequently debated questions, that associate the environmental elements to simple inputs that are necessary to productive processes, the consideration of the non-usual concepts of the traditional economic theory, like ecosystem services, resilience, entropy and hysteresis, would give better conditions to expand and relativize the interpretation of a series of premises and traditional economy dogmas. The natural capital concept, in this sense, represents itself as an essential and strategic concept, since it permits to accommodate all the complexity inherent to the environmental dimension and associate it to the economic system, fitting and marking out the sustainability framework. Still, the present work starts from the definition that both of the systems environmental and economic are essentially complex and that the effects of the relations between them are not trivial and have high levels of uncertainty associated to its dynamic. Whitin this orientation, this work proposed to realize a reflection about sustainability under the complex systems perspective, through a critical literature review and a multi-agent based modeling exercise, to simulate economic growth considering the environmental dimension. This analyses indicated that the incorporation of new concepts, from the complex systems theory, could establish a new support for the macroeconomic policies analysis, as well for the sustainability policies and, ultimately, to contribute to the strengthening of the basic ecological economy premises.

Patterns provide solutions to recurring design problems in a variety of domains, including that of software design. The best patterns are generative: they show how to build the solution they propose, rather than just explaining it. A collection of patterns that work together to generate a complex system is called a pattern language. Pattern languages have been written for domains as diverse as architecture and computer science, but the process of developing pattern languages is not well understood. This thesis focuses on defining both the structure of pattern languages and the processes by which they are built. The theoretical foundation of the work is existing theory on symmetry breaking. The form of the work is itself a pattern language: a Language Designer's Pattern Language (LDPL). LDPL itself articulates the structure of pattern languages and the key processes by which they form and evolve, and thus guides the building of a properly structured pattern language. LDPL uses multidisciplinary examples to validate the claims made, and an existing software pattern language is analyzed using the material developed. A key assumption of this thesis is that a pattern language is a structural entity; a pattern is not just a transformation on system structure, but also the resultant structural configuration. Another key assumption is that it is valid to treat a pattern language itself as a complex, designed system, and therefore valid to develop a pattern language for building pattern languages. One way of developing a pattern language for building pattern languages would be to search for underlying commonality across a variety of existing, well known pattern languages. Such underlying commonality would form the basis for patterns in LDPL. This project has not directly followed this approach, simply because very few pattern languages that are genuinely structural have currently been explicitly documented. Instead, given that pattern languages articulate structure and behavior of complex systems, this research has investigated existing complex systems theory - in particular, symmetry-breaking - and used that theory to underpin the pattern language. The patterns in the language are validated by examples of those patterns within two well known pattern languages, and within several existing systems whose pattern languages have not necessarily been explicitly documented as such, but the existence of which is assumed in the analysis. In addition to developing LDPL, this project has used LDPL to critique an existing software pattern language, and to show how that software pattern language could potentially have been generated using LDPL. Existing relationships between patterns in the software language have been analyzed and, in some cases, changes to patterns and their interconnections have been proposed as a way of improving the language. This project makes a number of key contributions to pattern language research. It provides a basis for semantic analysis of pattern languages and demonstrates the validity of using a pattern language to articulate the structure of pattern languages and the processes by which they are built. The project uses symmetry-breaking theory to analyze pattern languages and applies that theory to the development of a language. The resulting language, LDPL, provides language developers with a tool they can use to help build pattern languages.

The current research sets out to demonstrate a comprehensive approach to validate complex human performance models as applied to time-sensitive tasks. This document is divided into 4 sections. Section 1 (Chapters 1 – 3) outlines previous efforts in the literature that have attempted to validate complex human performance models in the field with an emphasis on manual control models, task network models, cognitive models and integrated architectures. Section 2 (Chapters 4 – 7) elaborates on a validation approach and applies it to a baseline model of a complex task in the air traffic control domain. Section 3 (Chapters 7-12) outlines the importance of adopting an iterative model development-model validation process and reports on the three model iterations in an attempt to improve the validity of the baseline model. Each model augmentation was validated using the same validation approach and measures that were defined in Section 2. Section 4 (Chapters 13-14) provides a discussion and interpretation of the model results and highlights contributions to the field of both model validation and the field of human performance modelling of complex systems.

The current research sets out to demonstrate a comprehensive approach to validate complex human performance models as applied to time-sensitive tasks. This document is divided into 4 sections. Section 1 (Chapters 1 – 3) outlines previous efforts in the literature that have attempted to validate complex human performance models in the field with an emphasis on manual control models, task network models, cognitive models and integrated architectures. Section 2 (Chapters 4 – 7) elaborates on a validation approach and applies it to a baseline model of a complex task in the air traffic control domain. Section 3 (Chapters 7-12) outlines the importance of adopting an iterative model development-model validation process and reports on the three model iterations in an attempt to improve the validity of the baseline model. Each model augmentation was validated using the same validation approach and measures that were defined in Section 2. Section 4 (Chapters 13-14) provides a discussion and interpretation of the model results and highlights contributions to the field of both model validation and the field of human performance modelling of complex systems.

Agriculture is undergoing extreme change. The introduction of new generation agricultural products has generated an increased need for efficient and accurate product segregation across a number of Canadian agricultural sectors. In particular, monitoring, controlling and preventing commingling of various wheat grades is critical to continued agri-food safety and quality assurance in the Canadian grain handling system. The Canadian grain handling industry is a vast regional supply chain with many participants. Grading of grain for blending had historically been accomplished by the method of Kernel Visual Distinguishability (KVD). KVD allowed a trained grain grader to distinguish the class of a registered variety of wheat solely by visual inspection. While KVD enabled rapid, dependable, and low-cost segregation of wheat into functionally different classes or quality types, it also put constraints on the development of novel traits in wheat. To facilitate the introduction of new classes of wheat to enable additional export sales in new markets, the federal government announced that KVD was to be eliminated from all primary classes of wheat as of August 1, 2008. As an alternative, the Canadian Grain Commission has implemented a system called Variety Eligibility Declaration (VED) to replace KVD. As a system based on self-declaration, the VED system may create moral hazard for misrepresentation. This system is problematic in that incentives exist for farmers to misrepresent their grain. Similarly, primary elevators have an incentive to commingle wheat classes in a profitable manner. Clearly, the VED system will only work as desired for the grain industry when supported by a credible monitoring system. That is, to ensure the security of the wheat supply chain, sampling and testing at some specific critical points along the supply chain is needed. While the current technology allows the identification of visually indistinguishable grain varieties with enough precision for most modern segregation requirements, this technology is relatively slow and expensive. With the potential costs of monitoring VED through the current wheat supply chain, there is a fundamental tradeoff confronting grain handlers, and effective handling strategies will be needed to maintain historical wheat uniformity and consistency while keeping monitoring costs down. There are important operational issues to efficiently testing grain within the supply chain, including the choice of the optimal location to test and how intensively to test. The testing protocols for grain deliveries as well as maintaining effective responsiveness to information feedback among farmers will certainly become a strategic emphasis for wheat handlers in the future. In light of this, my research attempts to identify the risks, incentives and costs associated with a functional declaration system. This research tests a series of incentives designed to generate truthful behavior within the new policy environment. In this manner, I examine potential and easy to implement testing strategies designed to maintain integrity and efficiency in this agricultural supply chain. This study is developed in the first instance by using an analytic model to explore the economic incentives for motivating farmer’s risk control efforts and handlers’ optimal handling strategies with respect to testing cost, penalty level, contamination risks and risk control efforts. We solve for optimal behavior in the supply chain assuming cost minimization among the participants, under several simplifying assumptions. In reality, the Canadian grain supply chain is composed of heterogeneous, boundedly rational and dynamically interacting individuals, and none of these characteristics fit the standard optimization framework used to solve these problems. Given this complex agent behavior, the grain supply chain is characterized by a set of non-linear relationships between individual participants, coupled with out of equilibrium dynamics, meaning that analytic solutions will not always identify or validate the set of optimized strategies that would evolve in the real world. To account for this inherent complexity, I develop an agent-based (farmers and elevators) model to simulate behaviour in a more realistic but virtual grain supply chain. After characterizing the basic analytics of the problem, the grain supply chain participants are represented as autonomous economic agents with a certain level of programmed behavioral heterogeneity. The agents interact via a set of heuristics governing their actions and decisions. The operation of a major portion of the Canadian grain handling system is simulated in this manner, moving from the individual farm up through to the country elevator level. My simulation results suggest testing strategies to alleviate misrepresentation (moral hazard) in this supply chain are more efficient for society when they are flexible and can be easily adjusted to react to situational change within the supply chain. While the idea of using software agents for modeling and understanding the dynamics of the supply chain under consideration is somewhat novel, I consider this exercise a first step to a broader modeling representation of modern agricultural supply chains. The agent-based simulation methodology developed in my dissertation can be extended to other economic systems or chains in order to examine risk management and control costs. These include food safety and quality assurance network systems as well as natural-resource management systems. Furthermore, to my knowledge there are no existing studies that develop and compare both analytic and agent-based simulation approaches for this type of complex economic situation. In the dissertation, I conduct explicit comparisons between the analytic and agent-based simulation solutions where applicable. While the two approaches generated somewhat different solutions, in many respects they led to similar overall conclusions regarding this particular agricultural policy issue.

Dinaminiai žmogaus organizmo kaip kompleksinės sistemos procesai atsispindi, registruojant fiziologinių procesų signalus, kuriems būdingos įvairaus laipsnio svyravimai, pavyzdžiui – elektrokardiogramos (EKG) signalai. Paprastai šių signalų analizei pasirenkami statistinės analizės metodai, kurie labiau taikytini globaliems organizmo procesams ir kuriems reikalingas didelis kiekis informacijos. Atitinkamai matematinės analizės metodai taikomi lokaliems organizmo procesams nagrinėti ir tuo pačiu nereikalingas toks didelis kiekis informacijos (mūsų atveju pakanka trijų atskaitymų). Fiziologinių signalų matricinė analizė padeda atskleisti sudėtingus žmogaus organizmo kaip KAS bei nuovargio fenomenus, parodančius netiesinių organizmo funkcinės būklės ir adaptacijos procesų fraktalinio ir chaotinio pobūdžio ypatumus. Atsižvelgiant į tai, šio tyrimo tikslas buvo nustatyti širdies ir kraujagyslių sistemos funkcinių rodiklių sąsajų kaitos ypatumus, vertinant sportuojančių asmenų organizmo būsenas. / The dynamic of the human body as a complex system of processes is reflected in the registration process of physiological signals which are characterized by varying degrees of oscillations, for example - an electrocardiogram (ECG) signals. Typically, these signals are selected for the statistical analysis methods that are more applicable to global processes of the body and requiring large amounts of information. Accordingly, the mathematical analysis methods are used to examine the local processes of the body and thus do not require such a large amount of information (in our case, enough of the three deductions). Physiological signal analysis matrix helps reveal the complex human organism as complex adaptive system and fatigue phenomena, which reflect the functional state of homeostasis and adaptation processes in fractal and chaotic nature of features. In this context, the aim of the study is to reveal the dynamical peculiarities of interactions of cardiovascular system indices in evaluation of sportsmen body’s functional state.

Submitted by Jos? Henrique Henrique (jose.neves@ufvjm.edu.br) on 2018-09-06T17:59:32Z No. of bitstreams: 2 license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) sophia_andrade_coelho.pdf: 1001228 bytes, checksum: 99ba4066b64da80f6d9ecaf0fd0d2b33 (MD5) / Approved for entry into archive by Rodrigo Martins Cruz (rodrigo.cruz@ufvjm.edu.br) on 2018-10-04T19:17:04Z (GMT) No. of bitstreams: 2 license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) sophia_andrade_coelho.pdf: 1001228 bytes, checksum: 99ba4066b64da80f6d9ecaf0fd0d2b33 (MD5) / Made available in DSpace on 2018-10-04T19:17:04Z (GMT). No. of bitstreams: 2 license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) sophia_andrade_coelho.pdf: 1001228 bytes, checksum: 99ba4066b64da80f6d9ecaf0fd0d2b33 (MD5) Previous issue date: 2018 / Sistemas a base de Multiagente (SMA) s?o formados por unidades aut?nomas de processamento de informa??o (agentes), munidos de arquitetura que lhes permite intera??o com seus semelhantes e seu ambiente ao longo do tempo. Tais sistemas podem ser desenvolvidos para o estudo de formas emergentes de linguagem e de intelig?ncia entre esses agentes. Por meio de tais modelos, no qual as vari?veis relacionadas ? comunica??o podem ser mais precisamente aferidas e controladas, pode-se estudar a origem e as interfer?ncias a respeito da linguagem natural, os fatores que a influenciam, bem como o que a pr?pria linguagem ? capaz de influenciar. A linguagem ? um dos fatores necess?rios para a abstra??o e intelig?ncia, e o estudo de sua emerg?ncia pode ser a pe?a chave para o cont?nuo desenvolvimento de mentes artificiais. As formas de intera??o e linguagem a emergir entre tais agentes beneficiam as capacidades de representa??o, e a comunica??o pode ser o ponto de partida para que os agentes desenvolvam e compartilhem conceitua??es do mundo, possibilitando o desenvolvimento de n?veis cognitivos elevados. Esta disserta??o tem como objetivo geral defender que a sem?ntica artificial existe e pode ser teoricamente sustentada a partir do pragmatismo e da teoria da ena??o, compreendendo o fen?meno da linguagem como um sistema complexo e din?mico emergido entre tais agentes. Os objetivos espec?ficos incluem uma an?lise e exposi??o das contribui??es de diversos campos das Ci?ncias Cognitivas, de forma a apontar a modelagem e simula??o Multiagente como m?todo capaz de rever, aprimorar e desenvolver teorias por um m?todo pass?vel de verifica??o. Como metodologia deste trabalho, adotou-se a revis?o bibliogr?fica por meio de an?lises, estudos e revis?es de peri?dicos, artigos, trabalhos e livros que abordaram intera??es complexas entre agentes, emerg?ncia da linguagem, semi?tica de Peirce e conceitos de lingu?stica cognitiva. Os resultados indicam que, compreendendo a linguagem como um sistema de rela??o simb?lica a partir do pragmatismo de Peirce e do m?todo anal?tico, os SMA s?o capazes de gerar formas de linguagem e consequentemente abstra??o e intelig?ncia. Assim, a linguagem ? abordada como um sistema complexo e multicausal que pode ser mais bem compreendido por meio de modelagens computacionais baseadas em sistemas complexos. A compet?ncia dos agentes de transformar o meio torna-os ativos na cria??o de formas pr?prias e genu?nas de linguagem e nas mudan?as de seu pr?prio sistema, formando um grupo funcional e inteligente. Conclui-se, com base na Lingu?stica Cognitiva e no pragmatismo, que a partir dos processos de intera??o a sem?ntica artificial existe, ? medida que pode-se gerar artificialmente jogos de linguagem e usos emergentes de signos como formas aut?nomas de representa??es de alto n?vel. / Disserta??o (Mestrado Profissional) ? Programa de P?s-Gradua??o em Ci?ncias Humanas, Universidade Federal dos Vales do Jequitinhonha e Mucuri, 2018. / The multi-agent based systems (MAS) are made up of information processing units (virtual agents), engineered to allow interaction among their peers and their own environment over time. These systems are developed for the study of the emergence of language and intelligence among agents. With these models, in which the variables related to communication can be more precisely measured and controlled, one can study the origin and the interferences that entail natural language. Also, the factors that influence language, and what language itself is able to influence could be better understood. Since language is one of the elements necessary for abstraction and intelligence, the study of its emergence may be the key element for the development of artificial intelligence. The forms of interaction and language to emerge among agents benefit capacities of representation so that communication may be the starting point for the agents to conceptualize their world, allowing them to develop high cognitive levels. Our general aim is to defend that artificial semantics exists from pragmatism and the theory of enation, understanding the phenomenon of language as a complex and dynamic system emerged among such agents. Our specific goal includes contributions to the advances in the fields of Artificial Intelligence and the reality of human language itself, in order to review, improve and develop theories by a selectable method. The methodology chosen was through research, studies, and reviews of periodicals, articles, works and books that involve complex interactions, language emergence, Peirce's semiotics and linguistic concepts. The results show that, comprehending language as a system of symbolic relations from pragmatism and analytic methods, MAS are able to generate forms of language and consequently abstraction and intelligence. This way, language is seen as a complex and multicausal system that may be better understood from computing modelling based in complex systems. The competence of the agents to transform their environment makes them active in creating and changings of their own system, forming a functional and intelligent group. Based on cognitive linguistics and pragmatism, we conclude that artificial semantics exist from interaction processes, since it is possible to notice the emergence of language and autonomous forms of high level representations.

abstract: The research on the topology and dynamics of complex networks is one of the most focused area in complex system science. The goals are to structure our understanding of the real-world social, economical, technological, and biological systems in the aspect of networks consisting a large number of interacting units and to develop corresponding detection, prediction, and control strategies. In this highly interdisciplinary field, my research mainly concentrates on universal estimation schemes, physical controllability, as well as mechanisms behind extreme events and cascading failure for complex networked systems. Revealing the underlying structure and dynamics of complex networked systems from observed data without of any specific prior information is of fundamental importance to science, engineering, and society. We articulate a Markov network based model, the sparse dynamical Boltzmann machine (SDBM), as a universal network structural estimator and dynamics approximator based on techniques including compressive sensing and K-means algorithm. It recovers the network structure of the original system and predicts its short-term or even long-term dynamical behavior for a large variety of representative dynamical processes on model and real-world complex networks. One of the most challenging problems in complex dynamical systems is to control complex networks. Upon finding that the energy required to approach a target state with reasonable precision is often unbearably large, and the energy of controlling a set of networks with similar structural properties follows a fat-tail distribution, we identify fundamental structural ``short boards'' that play a dominant role in the enormous energy and offer a theoretical interpretation for the fat-tail distribution and simple strategies to significantly reduce the energy. Extreme events and cascading failure, a type of collective behavior in complex networked systems, often have catastrophic consequences. Utilizing transportation and evolutionary game dynamics as prototypical settings, we investigate the emergence of extreme events in simplex complex networks, mobile ad-hoc networks and multi-layer interdependent networks. A striking resonance-like phenomenon and the emergence of global-scale cascading breakdown are discovered. We derive analytic theories to understand the mechanism of control at a quantitative level and articulate cost-effective control schemes to significantly suppress extreme events and the cascading process. / Dissertation/Thesis / Doctoral Dissertation Electrical Engineering 2016

abstract: A quantitative analysis of a system that has a complex reliability structure always involves considerable challenges. This dissertation mainly addresses uncertainty in- herent in complicated reliability structures that may cause unexpected and undesired results. The reliability structure uncertainty cannot be handled by the traditional relia- bility analysis tools such as Fault Tree and Reliability Block Diagram due to their deterministic Boolean logic. Therefore, I employ Bayesian network that provides a flexible modeling method for building a multivariate distribution. By representing a system reliability structure as a joint distribution, the uncertainty and correlations existing between system’s elements can effectively be modeled in a probabilistic man- ner. This dissertation focuses on analyzing system reliability for the entire system life cycle, particularly, production stage and early design stages. In production stage, the research investigates a system that is continuously mon- itored by on-board sensors. With modeling the complex reliability structure by Bayesian network integrated with various stochastic processes, I propose several methodologies that evaluate system reliability on real-time basis and optimize main- tenance schedules. In early design stages, the research aims to predict system reliability based on the current system design and to improve the design if necessary. The three main challenges in this research are: 1) the lack of field failure data, 2) the complex reliability structure and 3) how to effectively improve the design. To tackle the difficulties, I present several modeling approaches using Bayesian inference and nonparametric Bayesian network where the system is explicitly analyzed through the sensitivity analysis. In addition, this modeling approach is enhanced by incorporating a temporal dimension. However, the nonparametric Bayesian network approach generally accompanies with high computational efforts, especially, when a complex and large system is modeled. To alleviate this computational burden, I also suggest to building a surrogate model with quantile regression. In summary, this dissertation studies and explores the use of Bayesian network in analyzing complex systems. All proposed methodologies are demonstrated by case studies. / Dissertation/Thesis / Doctoral Dissertation Industrial Engineering 2018