"Dynamics" is very roughly the study of how objects change in time; for instance whether an electrical circuit goes to equilibrium, due to thermal dissipation. By "information", we mean how helpful it is to observe an object in order to know it better, for instance how many binary digits we can acquire on the value of a voltage by an appropriate measure. A "computation" is a physical process, e.g. the flow of current into a complex set of transistors, that after some time eventually gives us the solution of a mathematical problem (such as "Is 13 prime?"). We are interested to various relations between these concepts. In a first chapter, we unify some arguments in the literature to show that a whole class of quantities of dynamical systems are uncomputable. For instance the topological entropy of tilings and Turing machines. Then we propose a precise meaning to the statement "This dynamical system is a computer", at least for symbolic systems, such as cellular automata. We also show, for instance, that a "computer" must be dynamically unstable, and can even be chaotic. In a third chapter, we compare how complicated it is to control a system according whether we can acquire information on it ("feedback") or not ("open loop"). We are specifically interested in finite-state systems. In last chapter we show how to control a scalar linear system when only a finite amount of information can be acquired at every step of time.

Zwischen 1990 und 1994 wurden rund 1000 Liegenschaften, die in der ehemaligen DDR von der Sowjetarmee und der NVA für militärische Übungen genutzt wurden, an Bund und Länder übergeben. Die größten Truppenübungsplätze liegen in Brandenburg und sind heute teilweise in Großschutzgebiete integriert, andere Plätze werden von der Bundeswehr weiterhin aktiv genutzt. <br /> <br /> Aufgrund des militärischen Betriebs sind die Böden dieser Truppenübungsplätze oft durch Blindgänger, Munitionsreste, Treibstoff- und Schmierölreste bis hin zu chemischen Kampfstoffen belastet. Allerdings existieren auf fast allen Liegenschaften neben diesen durch Munition und militärische Übungen belasteten Bereichen auch naturschutzfachlich wertvolle Flächen; gerade in den Offenlandbereichen kann dies durchaus mit einer Belastung durch Kampfmittel einhergehen. Charakteristisch für diese offenen Flächen, zu denen u.a. Zwergstrauchheiden, Trockenrasen, wüstenähnliche Sandflächen und andere nährstoffarme baumlose Lebensräume gehören, sind Großflächigkeit, Abgeschiedenheit sowie ihre besondere Nutzung und Bewirtschaftung, d.h. die Abwesenheit von land- und forstwirtschaftlichem Betrieb sowie von Siedlungsflächen. Diese Charakteristik war die Grundlage für die Entwicklung einer speziell angepassten Flora und Fauna. <br /> <br /> Nach Beendigung des Militärbetriebs setzte dann in weiten Teilen eine großflächige Sukzession &ndash; die allmähliche Veränderung der Zusammensetzung von Pflanzen- und Tiergesellschaften &ndash; ein, die diese offenen Bereiche teilweise bereits in Wald verwandelte und somit verschwinden ließ. Dies wiederum führte zum Verlust der an diese Offenlandflächen gebundenen Tier- und Pflanzenarten.<br /> <br /> Zur Erhaltung, Gestaltung und Entwicklung dieser offenen Flächen wurden daher von einer interdisziplinären Gruppe von Naturwissenschaftlern verschiedene Methoden und Konzepte auf ihre jeweilige Wirksamkeit untersucht. So konnten schließlich die für die jeweiligen Standortbedingungen geeigneten Maßnahmen eingeleitet werden.<br /> <br /> Voraussetzung für die Einleitung der Maßnahmen sind zum einen Kenntnisse zu diesen jeweiligen Standortbedingungen, d.h. zum Ist-Zustand, sowie zur Entwicklung der Flächen, d.h. zur Dynamik. So kann eine Abschätzung über die zukünftige Flächenentwicklung getroffen werden, damit ein effizienter Maßnahmeneinsatz stattfinden kann.<br /> Geoinformationssysteme (GIS) spielen dabei eine entscheidende Rolle zur digitalen Dokumentation der Biotop- und Nutzungstypen, da sie die Möglichkeit bieten, raum- und zeitbezogene Geometrie- und Sachdaten in großen Mengen zu verarbeiten. Daher wurde ein fachspezifisches GIS für Truppenübungsplätze entwickelt und implementiert. Die Aufgaben umfassten die Konzeption der Datenbank und des Objektmodells sowie fachspezifischer Modellierungs-, Analyse- und Präsentationsfunktionen. Für die Integration von Fachdaten in die GIS-Datenbank wurde zudem ein Metadatenkatalog entwickelt, der in Form eines zusätzlichen GIS-Tools verfügbar ist. Die Basisdaten für das GIS wurden aus Fernerkundungsdaten, topographischen Karten sowie Geländekartierungen gewonnen.<br /> <br /> Als Instrument für die Abschätzung der zukünftigen Entwicklung wurde das Simulationstool AST4D entwickelt, in dem sowohl die Nutzung der (Raster-)Daten des GIS als Ausgangsdaten für die Simulationen als auch die Nutzung der Simulationsergebnisse im GIS möglich ist. Zudem können die Daten in AST4D raumbezogen visualisiert werden.<br /> Das mathematische Konstrukt für das Tool war ein so genannter Zellulärer Automat, mit dem die Flächenentwicklung unter verschiedenen Voraussetzungen simuliert werden kann. So war die Bildung verschiedener Szenarien möglich, d.h. die Simulation der Flächenentwicklung mit verschiedenen (bekannten) Eingangsparametern und den daraus resultierenden unterschiedlichen (unbekannten) Endzuständen. Vor der Durchführung einer der drei in AST4D möglichen Simulationsstufen können angepasst an das jeweilige Untersuchungsgebiet benutzerspezifische Festlegungen getroffen werden. / Between 1990 and 1994 more than a thousand military properties in East Germany used for military exercise by the Soviet army and the East German forces were handed over to government institutions of the national state and federal states (Laender), respectively. The largest military training areas are located in the federal state of Brandenburg. Some of these are, by now, part of environmentally protected areas, others are still being used by the German forces. <br /> <br /> As a result of long-lasting military operations the soils of these areas are heavily polluted by ammunition, duds, fuel and lubricants as well as a range of chemicals. Due to these special environmental conditions specifically adapted vegetation and fauna developed in these environments which make these areas valuable from an environmental protection perspective. Unlike other environments in North West Germany these former military training areas are characterised by wideness, separateness as well as the absence of agricultural, forest and settlement land use. This allowed for the development of a characteristic vegetation cover composed of dwarf bush heaths, drying lawns, desert-like sand surfaces and other nutrient-poor treeless habitats. <br /> <br /> Following the end of military use the process of succession &ndash; a gradual change of flora and fauna over space and time &ndash; set in. More a than a decade later the former open plains have, in part, developed into forests. The loss of open range land effected the disappearance of adapted animal and plant populations. To retain, preserve and develop these environmentally valuable areas different methods and concepts were examined for the respective effectiveness by multi-disciplinary group of scientists.<br /> <br /> Prerequisites of the application of methods to preserve the open plains are up-to-date knowledge of local conditions as well as in-depth understanding of the land surface developments and their dynamics, in particular. Based on this expertise the future surface development of these areas can be estimated which, in term, is a prerequisite for selecting the appropriate methods.<br /> <br /> Geo information systems (GIS) play an important role for the digital documentation and monitoring of biotopes, land use etc as they provide for acquisition, processing and storage of mass geo data including their locational and feature attributes. To make this GIS functionality available for the management and development of former military training areas a dedicated GIS has been developed and implemented. Tasks tackled included a database concept and object model as well as application-specific modelling, analysis and map presentation functions. Integration of relevant spatial and attribute data into this system required a meta data catalogue made available as an add-on GIS tool. <br /> Topographic base data for GIS input were taken from remote sensing imagery, topographic maps and additional field surveys.<br /> <br /> To simulate the development of the former military areas over space and time a specific simulation tool had to be developed as conventional GIS lack dynamic functionalities. The AST4D tool enables the user to utilise existing GIS data. Processed results, in term, can then be used within the GIS application. Additionally AST4D data may be visualised in map form.<br /> The AST4D GIS tool is based on the mathematical model of cellular automata. This model allows for the user-defined setting of status quo and simulation of future developments under different input parameters and target conditions.

Reversible circuits are similar to conventional logic circuits except that they are built from reversible gates. In reversible gates, there is a unique, one-to-one mapping between the inputs and outputs, not the case with conventional logic. Also, reversible gates require constant ancilla inputs for reconfiguration of gate functions and garbage outputs that help in keeping reversibility. Reversible circuits hold promise in futuristic computing technologies like quantum computing, quantum dot cellular automata, DNA computing, optical computing, etc. Thus, it is important to minimize parameters such as ancilla and garbage bits, quantum cost and delay in the design of reversible circuits. The first contribution of this dissertation is the design of a new reversible gate namely the TR gate (Thapliyal-Ranganathan) which has the unique structure that makes it ideal for the realization of arithmetic circuits such as adders, subtractors and comparators, efficient in terms of the parameters such as ancilla and garbage bits, quantum cost and delay. The second contribution is the development of design methodologies and a synthesis framework to synthesize reversible data path functional units, such as binary and BCD adders, subtractors, adder-subtractors and binary comparators. The objective behind the proposed design methodologies is to synthesize arithmetic and logic functional units optimizing key metrics such as ancilla inputs, garbage outputs, quantum cost and delay. A library of reversible gates such as the Fredkin gate, the Toffoli gate, the TR gate, etc. was developed by coding in Verilog for use during synthesis. The third contribution of this dissertation is the set of methodologies for the design of reversible sequential circuits such as reversible latches, flip-flops and shift registers. The reversible designs of asynchronous set/reset D latch and the D flip-flop are attempted for the first time. It is shown that the designs are optimal in terms of number of garbage outputs while exploring the best possible values for quantum cost and delay. The other important contributions of this dissertation are the applications of reversible logic as well as a special class of reversible logic called conservative reversible logic towards concurrent (online) and offline testing of single as well as multiple faults in traditional and reversible nanoscale VLSI circuits, based on emerging nanotechnologies such as QCA, quantum computing, etc. Nanoelectronic devices tend to have high permanent and transient faults and thus are susceptible to high error rates. Specific contributions include (i) concurrently testable sequential circuits for molecular QCA based on reversible logic, (ii) concurrently testable QCA-based FPGA, (iii) design of self checking conservative logic gates for QCA, (iv) concurrent multiple error detection in emerging nanotechnologies using reversible logic, (v) two-vectors, all 0s and all 1s, testable reversible sequential circuits.

ABSTRACT DEVELOPMENT OF A MICRO-FABRICATION PROCESS SIMULATOR FOR MICRO-ELECTRO-MECHANICAL SYSTEMS (MEMS) Yildirim, Alper M.S, Department of Mechanical Engineering Supervisor: Asst. Prof. Dr. Melik D&ouml / len December 2005, 140 pages The aim of this study is to devise a computer simulation tool, which will speed-up the design of Micro-Electro-Mechanical Systems by providing the results of the micro-fabrication processes in advance. Anisotropic etching along with isotropic etching of silicon wafers are to be simulated in this environment. Similarly, additive processes like doping and material deposition could be simulated by means of a Cellular Automata based algorithm along with the use of OpenGL library functions. Equipped with an integrated mask design editor, complex mask patterns can be created by the software and the results are displayed by the Cellular Automata cells based on their spatial location and plane. The resultant etched shapes are in agreement with the experimental results both qualitatively and quantitatively. Keywords: Wet Etching, Anisotropic Etching, Doping, Cellular Automata, Micro-fabrication simulation, Material Deposition, Isotropic Etching, Dry Etching, Deep Reactive Ion Etching

Thesis (M.Geom.M.)--University of Melbourne, Dept. of Geomatics, 2007. / Typescript. Includes bibliographical references (leaves 100-104).

Orientador: Paulina Setti Riedel / Banca: Cláudia Maria de Almeida / Banca: Maria Lúcia Marques / Resumo: A obtenção de informação sobre as condições da rede dutoviária representa papel primordial na operação dos dutos, não só com o propósito de manter a eficiência operacional, mas também, para a minimização dos riscos associados a possíveis acidentes, tanto decorrentes dos processos naturais quanto de causas antrópicas relacionadas ao uso e cobertura da terra. Neste sentido, o emprego de modelos matemáticos em estudos ambientais gera importantes contribuições científicas ao planejamento ordenado de uma área, uma vez que ajudam a entender o impacto das mudanças no uso e cobertura da terra e a prever alterações futuras nos ecossistemas. A área de estudo compreende porções dos municípios de Duque de Caxias, Belford Roxo e Nova Iguaçu e tem por objetivo realizar experimentos de mudanças da cobertura da terra em médio prazo (10 anos) para a faixa de dutos ORBEL. A modelagem realizada neste trabalho foi desenvolvida no software Dinamica-EGO, utilizando como entrada mapas classificados de cobertura da terra, obtidos através de imagens do satélite Landsat 5/TM, e variáveis espaciais para explicar as mudanças ocorridas na cobertura da terra. Para a calibração do modelo, utilizou-se o método de pesos de evidência, disponível no Dinamica-EGO, primeiramente categorizando-se os mapas de variáveis contínuas e, posteriormente, calculando-se os pesos de evidência. Valores positivos favorecem determinada transição, enquanto valores negativos indicam baixa probabilidade de transição, e valores próximos a zero não exercem efeito. A validação do modelo se deu pelo método de similaridade fuzzy, com decaimento exponencial. Os resultados mostram que, para os dois períodos analisados (1987-1998 e 1998-2010), a taxa de mudança para a transição "vegetação arbórea para outros" é bem maior do que para... (Resumo completo, clicar acesso eletrônico abaixo) / Abstract: Reliable information about conditions of pipeline network represents a major role in the operation of pipelines, not only in order to maintain operational efficiency, but also to minimize the risks associated with possible accidents such as rupture of the ducts due to landslides or mudslides mass block, causing damage to humans and the environment. The use of mathematical models in environmental studies allows significant scientific contributions to the physical planning of an area, since they help to understand the impact of changes in land use and land cover and predict future trends of changes in ecosystems. The study area includes portions of the municipalities of Duque de Caxias, Nova Iguaçu, and Belford Roxo and aims to carry out experiments in land-cover changes in the medium-term (10 years) within the catchment area of ORBEL pipelines. The modeling performed in this work was developed using the software Dinamica-EGO, using as input classified land cover maps obtained from Landsat 5/TM satellite images as well as spatial variables to explain the observed changes in land cover. For the model calibration, we used the method of weights of evidence that is available at Dinamica-EGO. Initially, the maps of continuous variables were categorized according to discrete ranges, which were then employed for the calculation of the respective positive weights of evidence. Positive values foster certain transitions, while negative values indicate low probability of transition, and values close to zero have no effect on changes. Model validation was executed by means of the fuzzy similarity method using exponential decay. The results show that for the two study periods (1987-1998 and 1998-2010) the rate of change for the transition 'others to woody vegetation' is much higher than for the other transitions observed in the same... (Complete abstract click electronic access below) / Mestre

Les transmissions radio à longue portée et basse énergie ouvrent de nouveaux champs d'application pour les capteurs, en particulier pour la surveillance de l'environnement. Le protocole radio LoRa permet, par exemple, de connecter des capteurs à une distance pouvant aller jusqu'à dix kilomètres en ligne de visée. Cependant, la grande surface couverte amène plusieurs difficultés, telles que le placement spatial en regard de la topologie géographique, ou la variabilité de la latence des communications. Le positionnement dans I'environnement comporte également des contraintes liées à I'intérêt des points de mesure du phénomène physique. Les critères de conception de ces réseaux tranchent donc avec les méthodes existantes (disques) quand on s'attaque aux terrains complexes. Cette thèse décrit des techniques de simulation basées sur I'analyse géographique cellulaire pour calculer les couvertures radio à longue portée et déduire les caractéristiques radios dans ces situations. Comme la propagation radio n'est qu'un cas particulier de phénomènes physiques, on montre qu'une approche unifiée cellulaire permet de caractériser beaucoup de comportements physiques potentiels. Le cas des fortes pluies et des inondations est étudié. L'analyse de la géographie est réalisée en utilisant des outils de segmentation pour produire des systèmes cellulaires qui sont à leur tour traduits en code pour des calculs de haute performance. La thèse fournit des résultats d'expériences de terrain complexes pratiques en utilisant LoRa, permettant de qualifier l'exactitude de la simulation des couvertures, et les caractéristiques d'ordonnancement des communications. Nous produisons des tables de performance pour les simulations sur les unités de traitement graphique (GPUs) qui montrent que le choix d'une algorithmique parallèle est pertinent sur ces problèmes. / Long-range radio transmissions open new sensor application fields, in particular for environment monitoring. For example, the LoRa radio protocol enables to connect remote sensors at distance as long as ten kilometers in a line-of-sight. However, the large area covered also brings several difficulties, such as the placement of sensing devices in regard to topology in geography, or the variability of communication latency. Sensing the environment also carries constraints related to the inlerest of sensing points in relation with a physical phenomenon. Thus criteria for designs are evolving a lot from the existing methods, especially in complex terrains. This thesis describes simulation techniques based on geography analysis to compute long-range radio coverages and radio characteristics in these situations. As radio propagation is just a particular case of physical phenomena, it is shown how a unified approach also allows to characterize the behavior of potential physical risks. The case of heavy rainfall and flooding is investigated. Geography analysis is achieved using segmentation tools to produce cellular systems which are in turn translated into code for high-þerformance computations. The thesis provides results from practical complex terrain experiments using LoRa which confirm the accuracy of the simulation, and scheduling characteristics for sample networks. Performance tables are produced for these simulations on current Graphics Processing Units (GPUs).

This thesis is a contribution to the formalisation of the notion of an agent within the class of finite multivariate Markov chains. In accordance with the literature agents are are seen as entities that act, perceive, and are goaldirected. We present a new measure that can be used to identify entities (called i-entities). The intuition behind this is that entities are spatiotemporal patterns for which every part makes every other part more probable. The measure, complete local integration (CLI), is formally investigated within the more general setting of Bayesian networks. It is based on the specific local integration (SLI) which is measured with respect to a partition. CLI is the minimum value of SLI over all partitions. Upper bounds are constructively proven and a possible lower bound is proposed. We also prove a theorem that shows that completely locally integrated spatiotemporal patterns occur as blocks in specific partitions of the global trajectory. Conversely we can identify partitions of global trajectories for which every block is completely locally integrated. These global partitions are the finest partitions that achieve a SLI less or equal to their own SLI. We also establish the transformation behaviour of SLI under permutations of the nodes in the Bayesian network. We then go on to present three conditions on general definitions of entities. These are most prominently not fulfilled by sets of random variables i.e. the perception-action loop, which is often used to model agents, is too restrictive a setting. We instead propose that any general entity definition should in effect specify a subset of the set of all spatiotemporal patterns of a given multivariate Markov chain. Any such definition will then define what we call an entity set. The set of all completely locally integrated spatiotemporal patterns is one example of such a set. Importantly the perception-action loop also naturally induces such an entity set. We then propose formal definitions of actions and perceptions for arbitrary entity sets. We show that these are generalisations of notions defined for the perception-action loop by plugging the entity-set of the perception-action loop into our definitions. We also clearly state the properties that general entity-sets have but the perception-action loop entity set does not. This elucidates in what way we are generalising the perception-action loop. Finally we look at some very simple examples of bivariate Markov chains. We present the disintegration hierarchy, explain it via symmetries, and calculate the i-entities. Then we apply our definitions of perception and action to these i-entities.

[pt] Esta dissertação investiga e propõe um novo simulador de circuitos de Autômatos Celulares com Pontos Quânticos (QCA) e uma nova metodologia para a criação e otimização de circuitos lógicos, utilizando técnicas da inteligência computacional. Autômatos Celulares com Pontos Quânticos é uma nova tecnologia, na escala nanométrica, que tem chamado a atenção dos pesquisadores por ser uma alternativa à tecnologia CMOS, cujo limite físico de miniaturização será atingido nos próximos anos. QCA tem um grande potencial no desenvolvimento de circuitos com maior densidade espacial, maior velocidade, baixa dissipação e baixo consumo de energia. Ao contrário das tecnologias tradicionais, QCA não codifica a informação pelo fluxo de corrente elétrica, mas pela configuração das cargas elétricas no interior das células. A interação coulombiana entre as células garante o fluxo da informação. Apesar de simples, essas características fazem com que a arquitetura de circuitos de QCA se torne não trivial. Portanto, a criação de um simulador e de uma metodologia de elaboração e síntese automática de circuitos possibilitam aos cientistas uma melhor visualização de como esses dispositivos funcionam, acelerando o desenvolvimento desses sistemas na escala nanométrica. Para atingir o objetivo proposto, técnicas de inteligência computacional, tais como redes neurais do tipo Hopfield, para o desenvolvimento do simulador, e algoritmos genéticos, para a metodologia de criação e otimização dos circuitos, foram empregadas. Os resultados encontrados foram significativos, comprovando que as técnicas da inteligência computacional podem ser uma ferramenta estratégica para o rápido desenvolvimento da nanoeletrônica e da nanotecnologia em geral. / [en] This dissertation investigates and considers a new simulator of Quantum Dots Cellular Automata (QCA) Circuits and a new methodology for the synthesis and optimization of logical circuits, by means of Computational Intelligence. Quantum-dot Cellular Automata (QCA) is a new technology in the nanometric scale which has called attention from researchers as one alternative for the CMOS technology, which is reaching its physical limitation. QCA have a large potential in the development of circuits with high space density and low heat dissipation, and can allow the development of faster computers with lower power consumption. Differently from the conventional technologies, QCA do not codify information by means of electric current flow, but rather by the configuration of electrical charges in the interior of the cells. The Coulomb interaction between cells is responsible by the flow of information. Despite simple, these features become the design of logical devices into a non-trivial task. Therefore, the development of a simulator and a methodology of automatic synthesis of QCA circuits make possible to the scientist a better evaluation of how these circuits work, accelerating the development of these new systems in the nanometer scale. To reach the proposed target, Computational Intelligence techniques were used. The first results show that these techniques are capable of simulating efficiently and fast, synthesizing optimized circuits with a reduced number of cells. Such optimization reduces the possibility of failures and guarantees higher speed.

Desde os primórdios da humanidade, a descoberta do método de processamento cerebral do som, e consequentemente da música, fazem parte do imaginário humano. Portanto, as pesquisas relacionadas a este processo constituem um dos mais vastos campos de estudos das áreas de ciências. Dentre as inúmeras tentativas para compreensão do processamento biológico do som, o ser humano inventou o processo automático de composição musical, com o intuito de aferir a possibilidade da realização de composições musicais de qualidade sem a imposição sentimental, ou seja, apenas com a utilização das definições e estruturas de música existentes. Este procedimento automático de composição musical, também denominado música aleatória ou música do acaso, tem sido vastamente explorado ao longo dos séculos, já tendo sido utilizado por alguns dos grandes nomes do cenário musical, como por exemplo, Mozart. Os avanços nas áreas de engenharia e computação permitiram a evolução dos métodos utilizados para composição de música aleatória, tornando a aplicação de autômatos celulares uma alternativa viável para determinação da sequência de execução de notas musicais e outros itens utilizados durante a composição deste tipo de música. Esta dissertação propõe uma arquitetura para geração de música harmonizada a partir de intervalos melódicos determinados por autômatos celulares, implementada em hardware reconfigurável do tipo FPGA. A arquitetura proposta possui quatro tipos de autômatos celulares, desenvolvidos através dos modelos de vizinhança unidimensional de Wolfram, vizinhança bidimensional de Neumann, vizinhança bidimensional Moore e vizinhança tridimensional de Neumann, que podem ser combinados de 16 formas diferentes para geração de melodias. Os resultados do processamento realizado pela arquitetura proposta são melodias no formato .mid, compostas através da utilização de dois autômatos celulares, um para escolha das notas e outro para escolha dos instrumentos a serem emulados, de acordo com o protocolo MIDI. Para tal esta arquitetura é formada por três unidades principais, a unidade divisor de frequência, que é responsável pelo sincronismo das tarefas executadas pela arquitetura, a unidade de conjunto de autômatos celulares, que é responsável pelo controle e habilitação dos autômatos celulares, e a unidade máquina MIDI, que é responsável por organizar os resultados de cada iteração corrente dos autômatos celulares e convertê-los conforme a estrutura do protocolo MIDI, gerando-se assim o produto musical. A arquitetura proposta é parametrizável, de modo que a configuração dos dados que influenciam no produto musical gerado, como por exemplo, a definição dos conjuntos de regras para os autômatos celulares habilitados, fica a cargo do usuário, não havendo então limites para as combinações possíveis a serem realizadas na arquitetura. Para validação da funcionalidade e aplicabilidade da arquitetura proposta, alguns dos resultados obtidos foram apresentados e detalhados através do uso de técnicas de obtenção de informação musical. / Since the beginnings of mankind, questions about the method of sound processing in the brain, and hence the music, are part of the human imagination. Therefore, the researches related to this process constitute one of the largest fields of research in science. Among many attempts to understand the biological processing of sound, humans beings invented the automatic musical composition automatic process in order to verify the possibility of producing quality musical compositions, without the human interference, wich means, only by using the definitions and structures of existing pieces of music. This automatic music composition procedure, also known as random music, has been widely exploited over the past centuries, being used by greats musicians like Mozart. The advances in engineering and computing allowed the evolution of random music composition methods, making the application of cellular automata a viable alternative to determine the execution sequence of musical notes, as well as, other items used in this type of composition. This dissertation proposes a hardware architecture for random music composition using cellular automata to determine the melodic intervals order. A prototype has been implemented in FPGA. The proposed hardware architecture has four kind of cellular automata, modeled according to the one-dimensional Wolfram neighborhood, two-dimensional Neumann neighborhood, two-dimensional Moore neighborhood and three-dimensional Neumann neighborhood. These cellular automata can be combined in sixteen different ways for generating melodies. The processing result as conducted by the proposed architecture are melodies in the .mid format, through the use of two cellular automata, one for selecting notes and the others for choosing the instruments to be emulated, in accordance with the MIDI protocol. This hardware is composed by three main units, the frequency divider unit, which is responsible for the synchronizing of the tasks performed by the hardware architecture, the cellular automata set unit, which is responsible for the control and enabling of the cellular automata and the MIDI machine, which is responsible for organizing the iterations of the cellular automata and convert them to the MIDI protocol structure, generating the final musical composition. The proposed hardware architecture is parametrizable, so that the data settings that influence the final product, such as, the rules of the cellular automata, are informed by the user. There are no limits regarding the possible combinations to be performed in the hardware architecture. In order to validate the functionality and applicability of the proposed hardware, some results were presented and detailed through the use of techniques for musical information retrieval.