Academic literature on the topic 'Architectures à passage de messages'
Create a spot-on reference in APA, MLA, Chicago, Harvard, and other styles
Consult the lists of relevant articles, books, theses, conference reports, and other scholarly sources on the topic 'Architectures à passage de messages.'
Next to every source in the list of references, there is an 'Add to bibliography' button. Press on it, and we will generate automatically the bibliographic reference to the chosen work in the citation style you need: APA, MLA, Harvard, Chicago, Vancouver, etc.
You can also download the full text of the academic publication as pdf and read online its abstract whenever available in the metadata.
Journal articles on the topic "Architectures à passage de messages"
1
Pym, Anthony. "On the passage of transcendent messages." FORUM / Revue internationale d’interprétation et de traduction / International Journal of Interpretation and Translation 14, no. 1 (August 4, 2016): 123–43. http://dx.doi.org/10.1075/forum.14.1.07pym.
Full textAbstract:
Abstract Translation is one way texts are accorded transcendence, understood as material transfer from a site of utterance. Although frequently construed as a quality of texts or auctorial virtue, transcendence is enacted by receivers (including translators) pulling texts across time and space, transforming them accordingly. Study of a war-commemoration text attributed to Atatürk shows this happening in its transfer to Australia. The historical authorship of the text has been contested, and analysis of its various translations and interpretations reveals competing interests, strategic omissions, distributed intercultural agency, and inscriptions. However, the historians involved in the debate, in both Turkey and Australia, have not sufficiently considered translation analysis, which can find some justification for the questioned text. Further, an ethics of cross-cultural communication might question the translation as an appeal to resolution based not just on the commonness of human suffering but also the shared concealment of guilt.
2
Hobley, Janet. "Rites of passage, moving into management: practice architectures and professionalism." Research in Post-Compulsory Education 24, no. 4 (October 2, 2019): 424–38. http://dx.doi.org/10.1080/13596748.2019.1654686.
Full text
3
Orhan, Orhan, and Huseyin Goren. "Largely Scalable Wireless Network Formation Architectures for Internet of Things." International Research Journal of Electronics and Computer Engineering 3, no. 4 (December 29, 2017): 17. http://dx.doi.org/10.24178/irjece.2017.3.4.17.
Full textAbstract:
As number of mobile devices increases with Internet of Things (IoT) capability and power consumption is important in multi hop communication, there is need of considering network formation architectures inIoT networks.In this work power levels of nodes, urgency of messages to be transmitted are not taken into consideration, a general approach considered with all nodes with same power levels and no priority for messages to be sent.Another assumption is all nodes are on the same 2D plane. Multi hop wireless network architectures studied and scalable, best performing architecture which is hypercubic network architecture is highlighted for IoT networks. Especially for large number of nodes needs to be considered hypercubic architecture performs much better than mesh, tree and ring kind of architectures in terms of dilation and number of connections. The simulation results are based on a simulator developed on C++ program. The results are showed that Hypercubic architecture with logarithmic dilation is much better than other network types. Cube Connected Cycles (CCC)based network architecture (which is derivative of Hypercube) can be preferred, especially when a constant degree is needed, in communication technologies such as Bluetooth. As a future work, a network architecture study can be made which takes energy levels and urgency of messages to be sent.
4
BARRETT, R. F., C. T. VAUGHAN, S. D. HAMMOND, and D. ROWETH. "REDUCING THE BULK IN THE BULK SYNCHRONOUS PARALLEL MODEL." Parallel Processing Letters 23, no. 04 (December 2013): 1340010. http://dx.doi.org/10.1142/s0129626413400100.
Full textAbstract:
For over two decades the dominant means for enabling portable performance of computational science and engineering applications on parallel processing architectures has been the bulk-synchronous parallel programming (BSP) model. Code developers, motivated by performance considerations to minimize the number of messages transmitted, have typically pursued a strategy of aggregating message data into fewer, larger messages. Emerging and future high-performance architectures, especially those seen as targeting Exascale capabilities, provide motivation and capabilities for revisiting this approach. In this paper we explore alternative configurations within the context of a large-scale complex multi-physics application and a proxy that represents its behavior, presenting results that demonstrate some important advantages as the number of processors increases in scale.
5
Alasadi, Emad, and Hamed S. Al-Raweshidy. "SSED: Servers Under Software-Defined Network Architectures to Eliminate Discovery Messages." IEEE/ACM Transactions on Networking 26, no. 1 (February 2018): 104–17. http://dx.doi.org/10.1109/tnet.2017.2763131.
Full text
6
González-García, A. César, Benito Vilas-Estévez, Elías López-Romero, and Patricia Mañana-Borrazás. "Domesticating Light and Shadows in the Neolithic: The Dombate Passage Grave (A Coruña, Spain)." Cambridge Archaeological Journal 29, no. 2 (December 6, 2018): 327–43. http://dx.doi.org/10.1017/s0959774318000562.
Full textAbstract:
Research on the Neolithic monuments and dwellings of Atlantic Europe has shown that plays of light and colour were tools for the social and symbolic construction of the world. The integration of the architectures into the surrounding landscape and the incorporation of the surrounding landscape into the architectures were an essential part of this logic. In this context, recent research in the megalithic passage grave of Dombate has evidenced an unusual physical manifestation of sunlight, which interacts with the decorated back stone. The light-and-shadow phenomenon occurs at sunrise during the period of winter solstice. In this paper we discuss the particulars of this phenomenon and we argue that sunlight when it penetrates the passage and chamber at sunrise on these dates may have dictated how the art was located and applied to the structural stone. Such differentiation seems to have had important cultural and ritual significance and encoded/embedded meaning for the tomb builders and may have implications for the consideration of the symbolic dimension of similar architectures in Atlantic Europe.
7
Figueira, Jorge, and Bruno Gil. "Biographies of Power: Personalities and Architectures." Joelho Revista de Cultura Arquitectonica, no. 8 (December 26, 2017): 170–200. http://dx.doi.org/10.14195/1647-8681_8_11.
Full textAbstract:
This section displays the work produced by the students of the Department of Architecture of the University of Coimbra (DARQ, UC) for the courses History of Architecture III and History of Architecture IV in the academic year 2016-2017. The main theme for the practical works of both course units was Biographies of Power: Personalities and Architectures. The work was exhibited in the gallery of DARQ, UC in September 2017. The works presented for Biographies of Power: Personalities were focused on the passage of testimony between Louis Sullivan and F. L. Wright. While they crucially contributed to the identity of a modern and American architecture, they carried out a professional and personal relationship that was also guided by the measurement of forces. The work done by the students for Biographies of Power: Architectures resulted from an observation of a set of buildings that goes beyond their architectural and physical contours, or which precisely searches for their actual description, in the light of any ideology that is ultimately represented or even reinforced by its own architectural condition.
8
TAGSOLD, CHRISTIAN. "Modernity, space and national representation at the Tokyo Olympics 1964." Urban History 37, no. 2 (July 6, 2010): 289–300. http://dx.doi.org/10.1017/s0963926810000362.
Full textAbstract:
ABSTRACT:The 1964 Tokyo Olympics acted as a rite of passage for post-war Japan, symbolizing the modernization of the city and the country. This was reflected by the space and architecture of the venues. Urban development of Olympic cities has been scrutinized recently but the symbolic implications have been touched upon only in passing, most especially in Tokyo's case. This article will show how symbolic layers of architecture and space aimed at linking history and modernity while bypassing the highly problematic legacy of ultra-nationalism and World War II. An important hub for transmitting this message was the Meiji Shrine dedicated to the first emperor of modern Japan. The hallmark building of the 1964 Games, Kenzo Tange's National Gymnasium, interacted with the shrine by way of an architectonic axis connecting them. This contrasted with the different spatial styles evident at the 1960 Olympics in Rome and 1972 Olympics in Munich, which testified to their different relationships to the national past. While developing infrastructure such as canalization and traffic was very important for Tokyo, symbolic revitalization of the city's fabric was equally crucial.
9
Kemp, P. S. "Meta-analyses, Metrics and Motivation: Mixed Messages in the Fish Passage Debate." River Research and Applications 32, no. 10 (October 2, 2016): 2116–24. http://dx.doi.org/10.1002/rra.3082.
Full text
10
Womick, Jake, John Eckelkamp, Sam Luzzo, Sarah J. Ward, S. Glenn Baker, Alison Salamun, and Laura A. King. "Exposure to authoritarian values leads to lower positive affect, higher negative affect, and higher meaning in life." PLOS ONE 16, no. 9 (September 15, 2021): e0256759. http://dx.doi.org/10.1371/journal.pone.0256759.
Full textAbstract:
Five studies tested the effect of exposure to authoritarian values on positive affect (PA), negative affect (NA), and meaning in life (MIL). Study 1 (N = 1,053) showed that simply completing a measure of right-wing authoritarianism (vs. not) prior to rating MIL led to higher MIL. Preregistered Study 2 (N = 1,904) showed that reading speeches by real-world authoritarians (e.g., Adolf Hitler) led to lower PA, higher NA, and higher MIL than a control passage. In preregistered Studies 3 (N = 1,573) and 4 (N = 1,512), Americans read authoritarian, egalitarian, or control messages and rated mood, MIL, and evaluated the passages. Both studies showed that egalitarian messages led to better mood and authoritarian messages led to higher MIL. Study 5 (N = 148) directly replicated these results with Canadians. Aggregating across studies (N = 3,401), moderational analyses showed that meaning in life, post manipulation, was associated with more favorable evaluations of the authoritarian passage. In addition, PA was a stronger predictor of MIL in the egalitarian and control conditions than in the authoritarian condition. Further results showed no evidence that negative mood (or disagreement) spurred the boost in MIL. Implications and future directions are discussed.
Dissertations / Theses on the topic "Architectures à passage de messages"
1
Benkahla-Aktouf, Oum-el-Kheir. "Diagnostic des architectures parallèles à passage de messages." Grenoble INPG, 1997. http://www.theses.fr/1997INPG0063.
Full textAbstract:
Ce memoire a pour objet le diagnostic, c'est-a-dire la detection et la localisation des elements defectueux, dans les architectures mimd a passage de messages. En raison du nombre important d'elements dans ces architectures, des methodes de test distribue sont appropriees. Une voie interessante consiste a exploiter la possibilite qu'ont les processeurs de se tester mutuellement. Tout d'abord, nous avons etudie les performances des strategies de tests interprocesseurs. Un outil et un modele pour l'evaluation des algorithmes de diagnostic ont ete proposes. Le comportement des algorithmes de diagnostic distribue a ainsi pu etre etudie. Par la suite, nous avons propose une strategie de diagnostic qui tienne compte des contraintes propres a une architecture parallele a passage de messages, en particulier l'acces limite aux noeuds. Cette strategie, appelee diagnostic par l'hote, assure la determination de l'etat de la machine par une entite externe, l'hote. Ceci permet d'envisager des operations de reparation ou de remplacement des elements defectueux. En, outre, une nouvelle mesure de diagnosabilite, appelee diagnosabilite par l'hote, a ete proposee. Par ailleurs, trois algorithmes fondes sur l'approche du diagnostic par l'hote, ont ete proposes. Ces algorithmes different par la strategie de test et de transmission des resultats de tests mise en oeuvre. Leurs performances ont ete evaluees a partir d'une modelisation a base de reseaux de files d'attente, puis comparees a celles d'un algorithme de diagnostic classique. L'extension du diagnostic par l'hote aux architectures paralleles ou la communication est assuree par des routeurs a egalement ete abordee. Il en a resulte un modele fonctionnel pour le diagnostic de ce type d'architectures, et un algorithme de diagnostic par l'hote fonde sur ce modele. Enfin, une approche de diagnostic hierarchique a ete proposee pour permettre le diagnostic par l'hote des systemes paralleles comportant un nombre eleve de processeurs.
2
Germain, Cécile. "Étude des mécanismes de communication pour une machine massivement parallèle : mega." Paris 11, 1989. http://www.theses.fr/1989PA112385.
Full textAbstract:
Le projet de Machine pour l'Expérimentation des Grandes Architectures (MEGA) de l'équipe Architecture et conception des Circuits Intégrés du LRI comporte à la fois l'étude d'un modèle de représentation des connaissances, les Réseaux de Processus Dynamiques, et d'une architecture matérielle support, la machine MEGA. MEGA est une machine MIMD à passage de messages, destinée à supporter un modèle de communication de type Ether. Son objectif est d'offrir un parallélisme matériel extrêmement massif, qui pourrait aller jusqu'à 1 Million de Processeurs Elémentaires, chacun de ces PE intégrant dans un seul composant VLSI une unité centrale, une mémoire locale et un circuit de gestion des communications. Les mécanismes de communication représentent un des points les plus sensibles d'une telle architecture : ils supportent une très forte contrainte technologique du point de vue de la faisabilité ; mais leurs performances déterminent très largement celles du système tout entier, puisque le parallélisme matériel ne peut être exploité pleinement que si le rapport communication/calcul reste équilibré. L'étude de ces mécanismes de communication comprend trois aspects : le modèle de messagerie en tant qu'interface processeur/réseau, la topologie du réseau d’interconnexion et la stratégie de routage. Dans les trois cas, l'objectif du parallélisme extrêmement massif nous a conduits à proposer des solutions originales vis-à-vis de celles utilisées dans les machines à passage de messages de grain moyen comme les hyper cubes commerciaux. Le modèle de messagerie proposé offre une relation directe entre processus lorsqu'elle est possible, tout supportant la dynamicité des applications. La comparaison de topologies à densité d'interconnexion constante montre que les grilles de petites dimensions offrent de meilleures performances que les hypercubes ; la grille 3D réalise alors un bon compromis performances/faisabilité. Nous avons développé une nouvelle stratégie de routage, le Routage Forcé, qui permet une utilisation efficace des liens du réseau en limitant la contention par dispersion aléatoire des messages sans mobiliser de ressources de surface pour le tamponnement. Nous avons spécifié le circuit de routage et d'interface processeur/réseau. Il gère complètement le transit des messages dans le réseau sans intervention du processeur de calcul et implante le Routage ForcéThe Machine to Explore Giant Architectures (MEGA) project of the "Architecture et Conception des Circuits Intégrés" team of the LRI includes both the study of a logical model, the Dynamic Processes Networks, and of a hardware architecture, the MEGA machine. MEGA is a MIMD message-passing machine, intended to support an Ether communication model. Its target is very massive parallelism, up to 1 Million Elementary Processors. Each of these processors integrates in a single VLSI chip a processing unit, a local memory and a communication management unit. The communication design is one of the key features of such an architecture. It bears a very high technological constraint from a feasibility point of view, but it's performances determines the whole system ones. Therefore hardware parallelism may be actually useful only if the communication to computation ratio may be balanced. The communication design study includes three points of view : the communication model as an interface between software and hardware communication tools, the network topology and the routing strategy. In the three cases, the extreme parallelism target induced us to propose solutions which are original beside these in use in medium-grain multicomputers such as the commercial hypercubes. The communication model offers a direct interprocess relation when possible and allows however the processes graph dynamicity. The constant wiring density comparison for topologies shows that low dimensional meshes have better performances than hypercubes; the 3D mesh is then a good compromise between performances and feasibility constraints. We developed also a new routing strategy which we called Forced Routing. This strategy provides an efficient channels use. It limits contention by a randomized dispersion of messages and only needs very little buffering area. We achieved a logical specification for a routing and network/processor interface circuit. This circuit manages the messages without interfering with the processing unit activity and implements Forced Routing
3
Damez, Lionel. "Approche multi-processeurs homogènes sur System-on-Chip pour le traitement d'image." Phd thesis, Université Blaise Pascal - Clermont-Ferrand II, 2009. http://tel.archives-ouvertes.fr/tel-00724443.
Full textAbstract:
La conception de prototypes de systèmes de vision en temps réel embarqué est sujet à de multiples contraintes sévères et fortement contradictoires. Dans le cas de capteurs dits "intelligents", il est nécessaire de fournir une puissance de traitement suffisante pour exécuter les algorithmes à la cadence des capteurs d'images avec un dispositif de taille minimale et consommant peu d'énergie. La conception d'un système monopuce (ou SoC) et l'implantation d'algorithmes de plus en plus complexes pose problème si on veut l'associer avec une approche de prototypage rapide d'applications scientifiques. Afin de réduire de manière significative le temps et les différents coûts de conception, le procédé de conception est fortement automatisé. La conception matérielle est basée sur la dérivation d'un modèle d'architecture multiprocesseur générique de manière à répondre aux besoins de capacité de traitement et de communication spécifiques à l'application visée. Les principales étapes manuelles se réduisent au choix et au paramétrage des différents composants matériels synthétisables disponibles. La conception logicielle consiste en la parallélisation des algorithmes, qui est facilitée par l'homogénéité et la régularité de l'architecture de traitement parallèle et la possibilité d'employer des outils d'aide à la parallélisation. Avec l'approche de conception sont présentés les premiers éléments constitutifs qui permettent de la mettre en oeuvre.Ceux ci portent essentiellement sur les aspects de conception matérielle. L'approche proposée est illustrée par l'implantation d'un traitement de stabilisation temps réel vidéo sur technologie SoPC
4
Aktouf, Chouki, and Guy Mazaré. "Test fonctionnel et reconfiguration d'architectures massivement parallèles à passage de messages." Grenoble INPG, 1995. http://www.theses.fr/1995INPG0001.
Full textAbstract:
Cette thèse a pour object la validation des machines massivement parallèles à passage de messages. Une stratégie de validation hiérarchique est proposée, autour des trois étapes suivantes: un test de routage, un test de mémoire et un test distribué des processeurs. Cette stratégie de validation est complétée par une phase de reconfiguration statique de la machine qui permet son exploitation après validation. Cette stratégie a été appliquée à une machine massivement parallèle, appelée, «machine cellulaire». Le test de routage est basé sur la technique de «Scan périphérique» et la norme IEEE 1149. 1. Le test de mémoire a consisté à évaluer la possibilité d'appliquer les algorithmes classiques de test de mémoire aux mémoires distribuées noyées dans une architecture massivement parallèle. Le test des processeurs est basé sur une stratégie de diagnostic distribué à travers le réseau: cette stratégie est fondée sur un test mutuel des nœuds en se basant sur un algorithme dont l'évolution dépend de l'état du réseau. La génération du programme de test exécuté par chaque processeur a été étudiée en tenant compte des contraintes de la machine cellulaire. La phase de validation permettant de déterminer l'ensemble des liens et des nœuds défectueux, il s'agit alors de proposer une reconfiguration statique de la machine de telle sorte qu'elle puisse supporter des applications avec les seuls nœuds corrects. Des algorithmes de routage tolérants aux fautes ont été proposés et évalués.
5
Genaud, Stéphane. "Exécutions de programmes parallèles à passage de messages sur grille de calcul." Habilitation à diriger des recherches, Université Henri Poincaré - Nancy I, 2009. http://tel.archives-ouvertes.fr/tel-00440503.
Full textAbstract:
Le document présente une synthèse de travaux sur le déploiement, l'utilisation et les techniques de mise en oeuvre d'applications développées selon un modèle de programmation à passage de messages sur des grilles de calcul. La première partie décrit les performances observées sur la période 2002-2006 sur une plateforme à l'échelle de la France, ainsi que les gains obtenus par équilibrage de charge. La deuxième partie décrit un intergiciel nouveau baptisé P2P-MPI qui synthétise un ensemble de propositions pour améliorer la prise en charge de tels programmes à passage de messages.
6
Boisvert, Sébastien. "Assemblage adaptatif de génomes et de méta-génomes par passage de messages." Thesis, Université Laval, 2014. http://www.theses.ulaval.ca/2014/30598/30598.pdf.
Full textAbstract:
De manière générale, les procédés et processus produisent maintenant plus de données qu’un humain peut en assimiler. Les grosses données (Big Data), lorsque bien analysées, augmentent la compréhension des processus qui sont opérationnels à l’intérieur de systèmes et, en conséquence, encouragent leur amélioration. Analyser les séquences de l’acide désoxyribonucléique (ADN) permet de mieux comprendre les êtres vivants, en exploitant par exemple la biologie des systèmes. Les séquenceurs d’ADN à haut débit sont des instruments massivement parallèles et produisent beaucoup de données. Les infrastructures informatiques, comme les superordinateurs ou l’informatique infonuagique, sont aussi massivement parallèles de par leur nature distribuée. Par contre, les ordinateurs ne comprennent ni le français, ni l’anglais – il faut les programmer. Les systèmes logiciels pour analyser les données génomiques avec des superordinateurs doivent être aussi massivement parallèles. L’interface de passage de messages permet de créer de tels logiciels et une conception granulaire permet d’entrelacer la communication et le calcul à l’intérieur des processus d’un système de calcul. De tels systèmes produisent des résultats rapidement à partir de données. Ici, les logiciels RayPlatform, Ray (incluant les flux de travail appelé Ray Meta et Ray Communities) et Ray Cloud Browser sont présentés. L’application principale de cette famille de produits est l’assemblage et le profilage adaptatifs de génomes par passage de messages.Generally speaking, current processes – industrial, for direct-to-consumers, or researchrelated – yield far more data than humans can manage. Big Data is a trend of its own and concerns itself with the betterment of humankind through better understanding of processes and systems. To achieve that end, the mean is to leverage massive amounts of big data in order to better comprehend what they contain, mean, and imply. DNA sequencing is such a process and contributes to the discovery of knowledge in genetics and other fields. DNA sequencing instruments are parallel objects and output unprecedented volumes of data. Computer infrastructures, cloud and other means of computation open the door to the analysis of data stated above. However, they need to be programmed for they are not acquainted with natural languages. Massively parallel software must match the parallelism of supercomputers and other distributed computing systems before attempting to decipher big data. Message passing – and the message passing interface – allows one to create such tools, and a granular design of blueprints consolidate production of results. Herein, a line of products that includes RayPlatform, Ray (which includes workflows called Ray Meta and Ray Communities for metagenomics) and Ray Cloud Browser are presented. Its main application is scalable (adaptive) assembly and profiling of genomes using message passing.
7
Bosilca, George. "Contribution à l'étude des environnements d'exécution parallèle fortement dynamiques à passage de messages." Paris 11, 2004. http://www.theses.fr/2004PA112316.
Full textAbstract:
Le calcul haute performance sur des architectures parallèles utilise différentes approches en fonction du modèle de mémoire de l'architecture, du niveau d'abstraction du langage de programmation et de la nature de l'application. La première partie de cette thèse est consacrée à la présentation d'un modèle d'exécution original basé sur l'appel de procédure à distance ou RPC (Remote Procedure Call), appelé Out-of-order execution parallel Virtual Machine (OVM). Le substrat a été construit suivant trois directions : portabilité via un modèle unique de mémoire, équilibrage de charge en utilisant un support modulaire et haute performance grâce à plusieurs optimisations. Je présente ici l'architecture générale d'OVM et je démontre ses performances pour différents types d'applications parallèles-: régulières, nécessitant un équilibrage de charge et des contraintes temps-réel. Avec l'augmentation du nombre de processeurs, les déconnexions et les disparitions des nœuds deviennent des évènements courants. La seconde partie est dédiée à l'environnement MPICH-V. Cet environnement de tolérance automatique des fautes est basé sur un checkpoint non coordonnée et un logging de messages distribué. MPICH-V se base sur des mémoires de canal, des serveurs de checkpoint et sur de preuves théoriques de ses protocoles. Se concentrant sur les mémoires de canal, des résultats sont présentés et analysés, pour des applications parallèles non triviales. Les résultats expérimentaux démontrent la bonne scalabilité et un haut degré de tolérance aux fautesGlobal Computing platforms, large scale clusters and future TeraGRID systems gather thousands of nodes for computing parallel scientific applications. These parallel architectures currently uses different approaches depending on the hardware memory model of the architecture, the abstraction level of the programming environment and the nature of the application. On the first part, I introduce an original client-server execution model based on RPCs called out-of-order parallel virtual machine (OVM). OVM aims to provide three main features: portability through a unique memory model, load-balancing using a plug-in support and high performance provided by several optimizations. The main optimizations are: non-blocking RPCs, data-flow management, persistent and non-persistent data, dynamic scheduling and asynchronous global operations. I present OVM general architecture and demonstrate high performance for several types of parallel applications: regular, with load balancing needs and with real-time constraints. At this scale, node failures or disconnections are frequent events. This Volatility reduces the MTBF of the whole system in the range of hour or minutes. On the second part, I present MPICH-V, an automatic fault tolerant MPI environment based on uncoordinated checkpoint/rollback and distributed message logging. MPICH-V architecture relies on Channel Memories, Checkpoint servers and theoretically proven protocols to execute SPMD and Master-Worker MPI applications on volatile nodes. Focusing on the channel memory, I present and analyse the performances for non-trivial parallel applications. Experimental results demonstrate good scalability and high tolerance to faults
8
Dujardin, Éric. "Reseau d'interconnexion pour une machine a passage de messages specialisee dans le traitement d'images." Paris 6, 1994. http://www.theses.fr/1994PA066558.
Full textAbstract:
Le sujet de cette these est l'etude d'un reseau d'interconnexion pour une machine a passage de messages. Le domaine d'application est le traitement d'image. Nous montrons que les algorithmes en image peuvent-etre implemente efficacement avec la methode diviser pour regner et que le traitement applique aux images genere en general une reduction d'information. Dans un premier temps, nous utilisons ces conclusions pour proposer une nouvelle topologie adaptee au traitement d'image: l'hyper-pyramide. C'est une pyramide tronquee de ces etages superieures. Dans chaque niveau sont definies des cellules, une cellule est un ensemble de processeurs, ou sont effectues les fusions des solutions intermediaires. Nous comparons cette nouvelle topologie avec d'autres topologies existantes en mesurant la complexite algorithmique et la complexite de realisation. Dans un deuxieme temps, nous developpons un simulateur de machines paralleles a passage de messages. Il accepte differents algorithmes de routage, differentes topologies et techniques de commutation. On peut aussi definir le traitement realise par les processeurs. On obtient, alors, un modele de communication de machines paralleles. Par ces simulations, nous comparons la topologie en hyper-pyramide, avec la maille, la pyramide et l'hypercube en utilisant differents algorithmes de routage et modeles de communication. Enfin, nous decrivons, une specification comportementale d'un circuit integre et nous le synthetisons avec des outils de synthese. Ce circuit est la realisation d'un processeur de communication avec une topologie en hyper-pyramide pour des processeurs de traitement comme le transputer
9
Khoumsi, Khalid. "Optimisation des performances dans les réseaux de communication des machines parallèles à passage de messages." Grenoble INPG, 1996. http://www.theses.fr/1996INPG0202.
Full textAbstract:
Les performances du reseau de communication representent la principale limitation des machines paralleles a memoire distribuee. Cette these traite de l'etude et la realisation de reseaux d'interconnexion permettant l'amelioration des performances des communications dans les machines paralleles mimd a passage de messages. Plusieurs architectures sont proposees dans deux cas tres differents. Tout d'abord dans le cas d'une machine cellulaire dont le but est d'offrir une tres grande puissance de calcul en interconnectant un grand nombre de processeurs tres simples. Une premiere solution permet de repousser la limitation technologique que represente le nombre de plots par chip. Deux autres solutions sont utilisees dans le but de reduire la latence des communications distantes par modification de la topologie du reseau. L'evaluation des solutions est faite apres implementation complete du reseau d'interconnexion. Ensuite dans le cas d'une machine parallele formee a partir de stations de travail interconnectees a travers un reseau local utilisant des liens hauts debits (par ex. Atm). Les solutions proposees traitent de l'adaptation de l'atm au calcul parallele. Les points abordes sont la taille des messages, le mode de communication sans connexion, et le controle a l'acces. Un simulateur de reseau a ete developpe pour l'evaluation des solutions proposees. Dans les deux cas, des modeles analytiques sont developpes pour predire la latence du reseau et mieux analyser les resultats experimentaux
10
Moukrim, Aziz. "Génération automatique de codes parallèles et nouvelles heuristiques d'ordonnancement pour les machines à passage de messages." Clermont-Ferrand 2, 1995. http://www.theses.fr/1995CLF21755.
Full textAbstract:
L'objet de cette these est la generation automatique de codes paralleles pour machines a passage de messages, et l'elaboration de nouvelles heuristiques d'ordonnancement avec delais de communication. Dans la premiere partie nous proposons un systeme de generation automatique de codes paralleles pour machines a passage de messages, descartes//, decompose en trois sous-systemes: sum//, ergo//et cogito//selon un niveau d'abstraction croissant. Nous avons procede a la realisation de sum//dont le langage d'entree permet deja a l'utilisateur de ne plus se soucier ni de la declaration ni de la gestion des canaux de communication dans les machines a passage de messages grace a la notion de classe de taches et tache copiee. Ensuite, nous avons propose un processus de traduction du sous-syteme ergo//en trois phases: partitionnement de l'algorithme d'entree et generation de scenarios, placement des taches et routage des donnees developpes en deuxieme partie de la these, et reconstitution des processus et gestion de leur traces sur les differents processuers sollicites. Dans la deuxieme partie, nous avons elabore de nouvelles heuristiques de placement des taches prenant en compte l'architecture reelle de la machine et considerant en meme temps le probleme de routage des donnees. Le gain moyen, obtenu par comparaison avec l'heuristique mcp utilisee dans le systeme hypertool et suivie d'une procedure de routage, est tres significatif. Enfin, nous avons etudie le probleme de l'encadrement du nombre optimal de processeurs necessaires pour l'obtention d'un ordonnancement optimal. Nous avons calcule une borne superieure dans le cas des arborescences et propose une heuristique pour l'evaluer dans le cas general. Une etude specifique a ete egalement menee pour les graphes de precedence sct
Books on the topic "Architectures à passage de messages"
1
Grabow, Patricia Jean. Messages from God: Passage from Fear Armageddon or Renaissance. Hara Publishing Group, 2002.
Find full text
2
Balasse, Marie, Jean-Philippe Brugal, Yannicke Dauphin, Eva-Maria Geigl, Christine Oberlin, and Ina Reiche, eds. Messages d'os. Editions des archives contemporaines, 2015. http://dx.doi.org/10.17184/eac.9782813001641.
Full textAbstract:
Os et dents résistent au passage du temps et à la décomposition du corps. Ils deviennent alors des fossiles, témoins précieux de la vie du passé, des archives biologiques et environnementales, archives culturelles lorsqu’ils sont retrouvés dans un contexte archéologique... Ils sont porteurs d’autant de « messages » qui peuvent être décryptés par des approches archéométriques.
3
Grabow, Patricia Jean. Messages from God: Passage from Karma a Coming Release Fron Sin and Repentance. Hara Publishing Group, 2002.
Find full text
4
Shoemaker, Pamela. The Gatekeeping of Political Messages. Edited by Kate Kenski and Kathleen Hall Jamieson. Oxford University Press, 2016. http://dx.doi.org/10.1093/oxfordhb/9780199793471.013.42.
Full textAbstract:
Simply put, gatekeeping involves deciding which messages to send to others and how to shape them. Billions of events occur each day, many with political ramifications, but there are more events than the news media can cover or even know about. Political information competes with all other topics to pass through source and media gates. Politicians create forces that constrain or facilitate this passage, but making it through one gate is usually not sufficient. Multiple gates in source, media, and audience channels winnow messages until we have those few that become television news stories or blog posts. To explain this process, gatekeeping theory was proposed in the late 1940s by Kurt Lewin; it is one of the older theories applied to the study of mass communication.
5
Eileen, Denza. Duties of Third States. Oxford University Press, 2016. http://dx.doi.org/10.1093/law/9780198703969.003.0047.
Full textAbstract:
This chapter examines Article 40 of the Vienna Convention on Diplomatic Relations which provides an outline of the duties of third states during diplomatic missions. The Article states that if a diplomatic agent, together with his family, passes through or is in the territory of a third State, the third State shall accord them inviolability and such other immunities as may be required to ensure their transit or return. The Article also states that in similar circumstances, the third State shall not hinder passage through their territories. Also, third States shall accord to official correspondence and other official communications in transit, including messages in code or cipher, the same freedom and protection as is accorded by the receiving State. Lastly, the Article states that the obligations of third States written in this Article shall also apply to the persons mentioned, and to official communications and diplomatic bags, whose presence in the territory of the third State is due to force majeure.
6
Zeidel, Robert F. Robber Barons and Wretched Refuse. Cornell University Press, 2020. http://dx.doi.org/10.7591/cornell/9781501748318.001.0001.
Full textAbstract:
This book explores the connection between the so-called robber barons who led American big businesses during the Gilded Age and Progressive Era and the immigrants who composed many of their workforces. As the book argues, attribution of industrial-era class conflict to an “alien” presence supplements nativism—a sociocultural negativity toward foreign-born residents—as a reason for Americans' dislike and distrust of immigrants. And in the era of American industrialization, employers both relied on immigrants to meet their growing labor needs and blamed them for the frequently violent workplace contentions of the time. The book uncovers the connection of immigrants to radical “isms” that gave rise to widespread notions of alien subversives whose presence threatened America's domestic tranquility and the well-being of its residents. Employers, rather than looking at their own practices for causes of workplace conflict, wontedly attributed strikes and other unrest to aliens who either spread pernicious “foreign” doctrines or fell victim to their siren messages. These characterizations transcended nationality or ethnic group, applying at different times to all foreign-born workers. The book concludes that, ironically, stigmatizing immigrants as subversives contributed to the passage of the Quota Acts, which effectively stemmed the flow of wanted foreign workers. Post-war employers argued for preserving America's traditional open door, but the negativity that they had assigned to foreign workers contributed to its closing.
7
Bishop, Sarah C. Undocumented Storytellers. Oxford University Press, 2019. http://dx.doi.org/10.1093/oso/9780190917159.001.0001.
Full textAbstract:
By projecting their stories into the public arena, undocumented storytellers refute mainstream discourse, trade anonymous narratives for individuality, and reveal the determination of those who elsewhere have been vilified by stereotype and presupposition. Taking a holistic approach to the role of storytelling in the immigrant rights movement, Bishop chronicles the ways young people uncover their lack of legal status experientially—through interactions with parents, in attempts to pursue rites of passage reserved for citizens, and as audiences of political and popular media. She provides both theoretical and pragmatic contextualization as activist narrators recount the experiences that influenced their decisions to cultivate public voices. Undocumented Storytellers offers a critical exploration of the ways undocumented immigrants harness the power of storytelling as a means of self-actualization, to mitigate the fear and uncertainty of life without legal status, and to advocate immigration reform. The book draws from a mixed methodology of forty in-depth interviews with undocumented immigrants from eighteen unique nations of origin, critical-rhetorical ethnographies of immigrant rights events and protests, and narrative analysis of immigrant-produced digital media to interrogate the power and limitations of autobiographical narrative activism. Offering an unparalleled view into the ways immigrants’ stories appear online, this book illuminates the power and limitations of digital narrative strategies by detailing how undocumented storytellers reframe their messages when stories have unintended consequences. The resulting work provides broad insights into the role of strategic framing and autobiographical story-sharing in advocacy and social movements.
Book chapters on the topic "Architectures à passage de messages"
1
Davis-Floyd, Robbi E. "Birth Messages." In Birth as an American Rite of Passage, 73–153. University of California Press, 2004. http://dx.doi.org/10.1525/california/9780520229327.003.0004.
Full text
2
Davis-Floyd, Robbi E. "How the Messages Are Received." In Birth as an American Rite of Passage, 187–240. University of California Press, 2004. http://dx.doi.org/10.1525/california/9780520229327.003.0006.
Full text
3
Goldman, Jane. "“Messages of Peace”." In Virginia Woolf, Europe, and Peace, 11–34. Liverpool University Press, 2020. http://dx.doi.org/10.3828/liverpool/9781949979374.003.0002.
Full textAbstract:
This chapter uncouples the binary opposition between peace and war to consider a number of unbridled notions of peace through close attention to its deployment in a rich nexus of Woolf’s writings. Proposing a Woolfian peace more properly understood as otium (a classical Latin term commonly translated as peace) with a queer twist, this chapter argues that Woolf, in collaboration with the intersectional pacifism of the Bloomsbury Group, offers a way of thinking through what it might mean to live, work and write peacefully. Offering a number of close and attentive readings, including a genetic account of the 'messages of peace' passage that straddles sections 9 and 10 of 'Time Passes' in To the Lighthouse, and the traces of Catullus that can be found in The Waves, this chapter demonstrates the presence of a radical and louche, ancient and avant-garde otium at work in Woolf’s writing.
4
Dehn, Torben. "The megalithic construction process and the building of passage graves in Denmark." In The Megalithic Architectures of Europe, 59–68. Oxbow Books, 2015. http://dx.doi.org/10.2307/j.ctvh1dpw8.10.
Full text
5
Politis, I., T. Dagiuklas, M. Tsagkaropoulos, and S. Kotsopoulos. "Interworking Architectures of 3G and WLAN." In Encyclopedia of Mobile Computing and Commerce, 357–64. IGI Global, 2007. http://dx.doi.org/10.4018/978-1-59904-002-8.ch059.
Full textAbstract:
The complex and demanding communications needs of modern humans led recently to the deployment of the 3G/UMTS mobile data networks and the wireless LANs. The already established GSM/GPRS radio access technology can easily handle the voice and low-rate data traffic such as short messages (SMS); however, it is inadequate for the more challenging real-time multimedia exchanges that require higher data rates and ubiquitous connectivity. The UTRAN radio access technology provides wide area coverage and multimedia services up to 2Mbps, while the recently deployed WLANs offer radio access at hotspots such as offices, shopping areas, homes, and other Internet/intranet-connected networks, with very high data rates up to 54Mbps (IEEE 802.11g). Hence, there is a strong need to integrate WLANs and 3G access technologies, and to develop a heterogeneous network based on an all-IP infrastructure that will be capable to offer ubiquitous and seamless multimedia services at very broadband rates.
6
Azzouzi, Salma, Sara Hsaini, and My El Hassan Charaf. "A Synchronized Test Control Execution Model of Distributed Systems." In Research Anthology on Architectures, Frameworks, and Integration Strategies for Distributed and Cloud Computing, 1377–95. IGI Global, 2021. http://dx.doi.org/10.4018/978-1-7998-5339-8.ch067.
Full textAbstract:
Conformance testing may be seen as mean to execute an IUT (implementation under test) by carrying out test cases in order to observe whether the behavior of the IUT is conforming to its specifications. However, the development of distributed testing frameworks is more complex and the implementation of the parallel testing components (PTCs) should take into consideration the mechanisms and functions required to support interaction during PTC communication. In this article, the authors present another way to control the test execution of PTCs by introducing synchronization messages into the local test sequences. Then, they suggest an agent-based simulation to implement synchronized local test sequences and resolve the problem of control and synchronization.
7
Rabadi, Yousef E., and Joan Lu. "Building a Secured XML Real-Time Interactive Data Exchange Architecture." In Research Anthology on Architectures, Frameworks, and Integration Strategies for Distributed and Cloud Computing, 270–357. IGI Global, 2021. http://dx.doi.org/10.4018/978-1-7998-5339-8.ch013.
Full textAbstract:
TCP and UDP communication protocols are the most widely used transport methods for carrying out XML data messages between different services. XML data security is always a big concern especially when using internet cloud. Common XML encryption techniques encrypt part of private sections of the XML file as an entire block of text and apply these techniques directly on them. Man-in-the-Middle and Cryptanalysts can generate statistical information, tap, sniff, hack, inject and abuse XML data messages. The purpose of this study is to introduce architecture of new approach of exchanging XML data files between different Services in order to minimize the risk of any alteration, data loss, data abuse, data misuse of XML critical business data information during transmission; by implementing a vertical partitioning on XML files. Another aim is to create a virtual environment within internet cloud prior to data transmission in order to utilise the communication method and rise up the transmission performance along with resources utilisation and spreads the partitioned XML file (shredded) through several paths within multi agents that form a multipath virtual network. Virtualisation in cloud network infrastructure to take advantage of its scalability, operational efficiency, and control of data flow are considered in this architecture. A customized UDP Protocol in addition to a pack of modules in RIDX adds a reliable (Lossless) and Multicast data transmission to all nodes in a virtual cloud network. A comparative study has been made to measure the performance of the Real-time Interactive Data Exchange system (RIDX) using RIDX UDP protocol against standard TCP protocol. Starting from 4 nodes up to 10 nodes in the domain, the results showed an enhanced performance using RIDX architecture over the standard TCP protocol.
8
Srivastav, Manoj Kumar, Robin Singh Bhadoria, and Tarasankar Pramanik. "Integration of Multiple Cache Server Scheme for User-Based Fuzzy Logic in Content Delivery Networks." In Handbook of Research on Advanced Applications of Graph Theory in Modern Society, 386–96. IGI Global, 2020. http://dx.doi.org/10.4018/978-1-5225-9380-5.ch016.
Full textAbstract:
The internet plays important role in the modern society. With the passage of time, internet consumers are increasing. Therefore, the traffic loads during communication between client and its associated server are getting complex. Various networking systems are available to send the information or to receive messages via the internet. Some networking systems are so expensive that they cannot be used for the regular purpose. A user always tries to use that networking system that works on expansion of optimizing the cost. A content delivery network (CDN) also called as content distribution network has been developed to manage better performance between client and list of available servers. This chapter presents the mathematical model to find optimization among client and cache server during delivery of content based on fuzzy logic.
9
Webb, Thomas E. "R v Secretary of State for the Home Department, ex parte Brind [1991] UKHL 4, House of Lords." In Essential Cases: Public Law. Oxford University Press, 2020. http://dx.doi.org/10.1093/he/9780191897689.003.0037.
Full textAbstract:
Essential Cases: Public Law provides a bridge between course textbooks and key case judgments. This case document summarizes the facts and decision in R v Secretary of State for the Home Department, ex parte Brind [1991] UKHL 4, House of Lords. The case considered whether the Secretary of State could restrict the editorial decisions of broadcasters as regards the way in which messages from spokespersons for proscribed organizations were broadcast. The United Kingdom was a signatory to the European Convention on Human Rights (ECHR) when the case was heard, but the case also predates the passage of the Human Rights Act 1998. There is discussion of the legal position of the ECHR under the common law in the United Kingdom, and the concept of proportionality in United Kingdom jurisprudence. The document also includes supporting commentary from author Thomas Webb.
10
Webb, Thomas E. "R v Secretary of State for the Home Department, ex parte Brind [1991] UKHL 4, House of Lords." In Essential Cases: Public Law. Oxford University Press, 2021. http://dx.doi.org/10.1093/he/9780191926440.003.0037.
Full textAbstract:
Essential Cases: Public Law provides a bridge between course textbooks and key case judgments. This case document summarizes the facts and decision in R v Secretary of State for the Home Department, ex parte Brind [1991] UKHL 4, House of Lords. The case considered whether the Secretary of State could restrict the editorial decisions of broadcasters as regards the way in which messages from spokespersons for proscribed organizations were broadcast. The United Kingdom was a signatory to the European Convention on Human Rights (ECHR) when the case was heard, but the case also predates the passage of the Human Rights Act 1998. There is discussion of the legal position of the ECHR under the common law in the United Kingdom, and the concept of proportionality in United Kingdom’s domestic jurisprudence. The document also includes supporting commentary from author Thomas Webb.
Conference papers on the topic "Architectures à passage de messages"
1
Karam, Ralph, Michel Salomon, and Raphael Couturier. "A Comparative Study of Deep Learning Architectures for Detection of Anomalous ADS-B Messages*." In 2020 7th International Conference on Control, Decision and Information Technologies (CoDIT). IEEE, 2020. http://dx.doi.org/10.1109/codit49905.2020.9263880.
Full text
2
Liu, Hui, Zhan Shi, Jia-Chen Gu, Quan Liu, Si Wei, and Xiaodan Zhu. "End-to-End Transition-Based Online Dialogue Disentanglement." In Twenty-Ninth International Joint Conference on Artificial Intelligence and Seventeenth Pacific Rim International Conference on Artificial Intelligence {IJCAI-PRICAI-20}. California: International Joint Conferences on Artificial Intelligence Organization, 2020. http://dx.doi.org/10.24963/ijcai.2020/535.
Full textAbstract:
Dialogue disentanglement aims to separate intermingled messages into detached sessions. The existing research focuses on two-step architectures, in which a model first retrieves the relationships between two messages and then divides the message stream into separate clusters. Almost all existing work puts significant efforts on selecting features for message-pair classification and clustering, while ignoring the semantic coherence within each session. In this paper, we introduce the first end-to- end transition-based model for online dialogue disentanglement. Our model captures the sequential information of each session as the online algorithm proceeds on processing a dialogue. The coherence in a session is hence modeled when messages are sequentially added into their best-matching sessions. Meanwhile, the research field still lacks data for studying end-to-end dialogue disentanglement, so we construct a large-scale dataset by extracting coherent dialogues from online movie scripts. We evaluate our model on both the dataset we developed and the publicly available Ubuntu IRC dataset [Kummerfeld et al., 2019]. The results show that our model significantly outperforms the existing algorithms. Further experiments demonstrate that our model better captures the sequential semantics and obtains more coherent disentangled sessions.
3
Козаченко, Екатерина, Ekaterina Kozachenko, Евгения Давыдова, and Evgenia Davydova. "A "Labyrinth" toy for 6-7 Years Old Children as a Way of Intellectual Development of a Child." In 29th International Conference on Computer Graphics, Image Processing and Computer Vision, Visualization Systems and the Virtual Environment GraphiCon'2019. Bryansk State Technical University, 2019. http://dx.doi.org/10.30987/graphicon-2019-2-202-206.
Full textAbstract:
The article considers the problem of a child’s quick loss of interest in a toy and its solution by creating an intellectual toy based on the experience of scientific research in the field of labyrinthology. Passing puzzles in the form of a maze have a beneficial effect on the child’s intellect, develops spatial thinking, memory and teaches the child how to find solutions to problems at a subconscious level, also disciplines and develops concentration. For a hull development the art design method was applied. A prototype of the designed toy is a cryptex - a portable storage used to hide secret messages invented by Leonardo Da Vinci and implemented by Justin Kirk Nevinson. The labyrinth system is based on the fractal tree method. The fractal labyrinth has a hierarchical (grid) structure, which creates more chances for a quick passage of the labyrinth without obstacles, but since the goal of the toy is to develop the intelligence of the child, it was decided to complicate the structure of the labyrinth by adding dead-end branches and additional tracks. Entering a dead-end branch allows the child to find and analyze the optimal moves in the game.
4
Inhestern, Lukas Benjamin, James Braun, Guillermo Paniagua, and José Ramón Serrano Cruz. "Design, Optimization and Analysis of Supersonic Radial Turbines." In ASME Turbo Expo 2019: Turbomachinery Technical Conference and Exposition. American Society of Mechanical Engineers, 2019. http://dx.doi.org/10.1115/gt2019-91756.
Full textAbstract:
Abstract New compact engine architectures such as pressure gain combustion require ad-hoc turbomachinery to ensure an adequate range of operation with high performance. A critical factor for supersonic turbines is to ensure the starting of the flow passages, which limits the flow turning and airfoil thickness. Radial outflow turbines inherently increase the cross section along the flow path, which holds great potential for high turning of supersonic flow with a low stage number and guarantees a compact design. First the preliminary design space is described. Afterwards a differential evolution multi-objective optimization with 12 geometrical design parameters is deducted. With the design tool AutoBlade 10.1, 768 geometries were generated and hub, shroud, and blade camber line were designed by means of Bezier curves. Outlet radius, passage height, and axial location of the outlet were design variables as well. Structured meshes with around 3.7 million cells per passage were generated. Steady three dimensional Reynolds averaged Navier Stokes (RANS) simulations, enclosed by the k-omega SST turbulence model were solved by the commercial solver CFD++. The geometry was optimized towards low entropy and high power output. To prove the functionality of the new turbine concept and optimization, a full wheel unsteady RANS simulation of the optimized geometry exposed to a nozzled rotating detonation combustor (RDC) has been performed and the advantageous flow patterns of the optimization were also observed during transient operation.
5
Quarta, Aurora. "Il Castello di “Carta”. Excursus della presenza del castello di Gallipoli nella cartografia storica." In FORTMED2020 - Defensive Architecture of the Mediterranean. Valencia: Universitat Politàcnica de València, 2020. http://dx.doi.org/10.4995/fortmed2020.2020.11339.
Full textAbstract:
The Castle of "paper". Excursus of Gallipoli’s castle presence in historical cartographyThe castle is located at the eastern part of the Gallipoli’s old town: the first data in archives and libraries started from the sixth century under the mention of castrum and in the following centuries there are many informations on parchments, written documents and bibliography published until today. The Syllabus Grecarum Membranarum from the twelfth century and the Statutum de reparatione castrorum of Frederick II are two precious sources about the primitive castle’s architecture.The structure endured the passage of the Byzantines, Normans, Swabians, Angevins and again, Aragonese, Venetians, Spaniards, Austrians and finally the Bourbons, until it became property of the State and now of the Gallipoli’s municipality. It has suffered over time numerous interventions to adapt it to new military needs: the castle was no longer effective with leading defence from new siege weapons, as for other architectures of the same period.The numerous representations preserved in Italian and European archives give a complete picture of the Gallipoli’s urban development and include the defensive system of the city: the different views illustrate the walls and allow us to understand the castle’s main evolutionary dynamics and its connection with the town.
6
Bettocchi, R., M. Pinelli, P. R. Spina, and M. Venturini. "Artificial Intelligence for the Diagnostics of Gas Turbines: Part II — Neuro-Fuzzy Approach." In ASME Turbo Expo 2005: Power for Land, Sea, and Air. ASMEDC, 2005. http://dx.doi.org/10.1115/gt2005-68027.
Full textAbstract:
In the paper, Neuro-Fuzzy Systems (NFSs) for gas turbine diagnostics are studied and developed. The same procedure used previously for the set up of Neural Network (NN) models was used. In particular, the same database of patterns was used for both training and testing the NFSs. This database was obtained by running a Cycle Program, calibrated on a 255 MW single shaft gas turbine working in the ENEL combined cycle power plant of La Spezia (Italy). The database contains the variations of the Health Indices (which are the characteristic parameters that are indices of gas turbine health state, such as efficiencies and characteristic flow passage areas of compressor and turbine) and the corresponding variations of the measured quantities with respect to the values in new and clean conditions. The analyses carried out are aimed at the selection of the most appropriate NFS structure for gas turbine diagnostics, in terms of computational time of the NFS training phase, accuracy and robustness towards measurement uncertainty during simulations. In particular, Adaptive Neuro-Fuzzy Inference System (ANFIS) architectures were considered and tested, and their performance was compared to that obtainable by using the NN models. An analysis was also performed in order to identify the most significant ANFIS inputs. The results obtained show that ANFISs are robust with respect to measurement uncertainty, and, in all the cases analyzed, the performance (in terms of accuracy during simulations and time spent for the training phase) proved to be better than that obtainable by MIMO and MISO Neural Networks trained and tested on the same data.