No description available.

No description available.

Developing syntactic complexity in writing is an important goal for many adult language learners. However, little is known about the effect of different writing practice tasks on the development of syntactic complexity. Furthermore, it is unclear to what degree syntactic complexity is related to writing quality. The main purpose of this study was to compare the longitudinal effects of three types of writing tasks on the development of syntactic complexity and writing quality. In order to investigate these relationships, a mixed-methods design was used. A quantitative component that was a quasi-experimental, longitudinal investigation was combined with an embedded qualitative component which involved eliciting stimulated recalls from participants from two of the experimental groups in the study. For the quantitative component, the participants (N = 105) were first-year, non-English majors at a four-year, co-educational university in western Japan. The participants were randomly divided into two groups: a sentence-combining group and a translation group. These two experimental groups were compared to an intact timed-writing group, which served as a comparison group. Writing samples were collected at three points throughout the year and analyzed based on five measures of syntactic complexity: mean length of sentence, mean length of T-unit, mean length of clause, clauses per T-unit, and T-units per sentence. Repeated-measures ANOVAs were used to analyze the five measures to investigate statistical differences across time for each group. One-way ANOVAs were used to analyze the five measures to ascertain differences between groups at each time. Furthermore, all the writing samples were evaluated by human raters for writing quality using an analytic rubric. Ratings were analyzed and investigated for changes in quality across time and between groups, as well as to examine the overall relationship between syntactic complexity and quality. For the qualitative component, seven participants were selected from the experimental groups to participate in stimulated recalls. The data were analyzed and coded to investigate the cognitive processes underlying each task. The results of the quantitative study indicated that the participants in the translation and sentence-combining groups made small, but significant gains in syntactic complexity across time. The translation group made significant growth on clauses per T-unit while the sentence-combining group made significant growth across time on mean length of sentence, mean length of T-unit, and T-units per sentence. The timed-writing group was not able to sustain significant growth across time. In the between-groups analyses, there were significant differences between the sentence-combining group and the other groups on mean length of sentence and T-units per sentence, indicating that the participants in that group demonstrated more coordination, and to a lesser degree, more subordination in their writing. In terms of quality of writing, the results indicated an overall weak, statistically significant positive correlation between measures of syntactic complexity and human rating judgments. Among all of the syntactic measures, T-units per sentence correlated the strongest with rating scores. The results of the qualitative component indicated that the participants practicing sentence-combining tasks focused most of their attention toward syntactic aspects of language while the participants in the translation group focused their attention on lexical aspects of language while performing the tasks. The findings of this study show that the development of syntactic complexity is influenced by writing practice tasks and that tasks that direct learners’ attention toward features of the language are more likely to hasten the development of syntactic complexity. Among the tasks, sentence-combining showed the most potential in developing syntactic complexity, particularly for increasing the use of compound sentences. However, none of the writing practice tasks led to significant gains in writing quality, in part because syntactic complexity was shown to have a moderately weak relationship to overall writing quality. / Teaching & Learning

Recently, multi-priced timed automata have received much attention for real-time systems. These automata extend priced timed automata by featuring several price parameters. This permits to compute objectives like the optimal ratio between rewards and costs. Arising from the model of timed automata, the multi-weighted setting has also attracted much notice for classical nondeterministic automata. The present thesis develops multi-weighted MSO-logics on finite, infinite and timed words which are expressively equivalent to multi-weighted automata, and studies decision problems for them. In addition, a Nivat-like theorem for weighted timed automata is proved; this theorem establishes a connection between quantitative and qualitative behaviors of timed automata. Moreover, a logical characterization of timed pushdown automata is given.

Under senare år har vi med hjälp av robotar som använder rörelseplanering kunnat automatisera olika processer och uppgifter. Idag finns det väldigt få strategier för lokal rörelseplanering vid jämförelse med global rörelseplanering. Syftet med det här projektet har varit att analysera tre strategier för lokal rörelseplanering, dessa har varit Dynamic Window Approach (DWA), Elastic Band (Eband) och Timed Elastic Band (TEB).I projektet har styrkor, svagheter, beteenden och förbättringsmöjligheter för respektive strategi studerats närmare genom att utföra olika simulerade tester. I testerna mätes tid för att nå mål, antal kollisioner och antalet gånger som målet nåddes. Under projektet användes en virtuell allriktad robot från ABB för att testa strategierna. Testerna genomfördes på ett så rättvist sätt som möjligt, där alla strategier fick samma antal försök och hade samma information om robotens begränsningar.Resultatet visar att TEB är den snabbaste strategin, följt av DWA och sista Eband som var den långsammaste strategin. TEB var också den strategi som presterade bäst vid dynamiska hinder, däremot var den också den strategi som kolliderade mest i testerna, medan Eband kolliderade minst. / In recent years, we have been able to automate various processes and tasks with the help of robots that use motion planning. Today, there are very few strategies for local motion planning when compared to global motion planning. The purpose of this project has been to analyze three strategies for local motion planning, these have been Dynamic Window Approach (DWA), Elastic Band (Eband) and Timed Elastic Band (TEB).In the project, strengths, weaknesses, behaviours and opportunities for improvement for each strategy have been studied in more detail by performing various simulated tests. The tests measure time to reach the goal, the number of collisions and the number of succeeding attempts. During the project, a virtual omni-directional robot from ABB was used to perform the tests. The tests were performed in as fair a way as possible, where all strategies got the same number of attempts and had the same information about the robot's limitations.The results show that TEB is the fastest strategy, followed by DWA and last Eband that was the slowest strategy. TEB was also the strategy that performed best in dynamic obstacles, however, it was also the strategy that collided most of the tests, while Eband collided the least.

La plupart des scénarios multimédia interactifs sont basés sur des spécifications informelles, il n'est donc pas possible de vérifier formellement des propriétés de ces systèmes. Nous préconisons la nécessité d'un modèle général et formel. Partitions interactives est un formalisme pour décrire des scénarios multimédia interactifs. Nous proposons une nouvelle sémantique pour les partitions interactives basée sur les structures d'événements temporisés. Avec une telle sémantique, nous pouvons spécifier des propriétés pour le système, en particulier, des propriétés sur les traces, qui sont difficiles à préciser avec la programmation par contraintes. Nous présentons également une sémantique opérationnelle des partitions interactives basée sur le calcul non-déterministe, temporisé, concurrent, par contraintes (ntcc) et nous rapportons la sémantique operationelle à la semantique en structures d'événements temporisés. Avec la sémantique opérationnelle, nous pouvons décrire formellement le comportement d'un scenario dont les durées des objets temporels peuvent être des intervalles d'entiers arbitraires. La sémantique opérationnelle est obtenue à partir de la sémantique en structures d'événements temporisés de la partition interactive. Pour fournir une telle traduction, nous avons d'abord défini la forme normale d'une structure d'événements temporisés, dans laquel les événements liés avec une durée zéro sont regroupés en un seul. Nous avons également défini la notion de structures d'événements temporisés répartissables, de telle sorte que son graphe de contraintes peut être expédié en se fondant uniquementsur la propagation locale. Nous croyons que la sémantique opérationnelle basée sur ntcc offre certains avantages par rapport à la sémantique des partitions interactives basée sur des réseaux de Petri; par exemple, les durées des objets temporels peuvent être des intervalles d'entiers arbitraires, tandis que dans la plupart des modèles de partitions interactives, les intervalles ne peut être utilisés que pour représenterles relations telles que l'égalité et les inégalités. Nos modèles ntcc de partitions interactives sont exécutés en utilisant Ntccrt, un interprète temps réel pour ntcc. Nos modèles peuvent également être vérifiés automatiquement en utilisant ntccMC, un vérificateur pour ntcc, de temps borné, basée sur les automates finis, que nous introduisons dans cette thèse. En utilisant ntccMC, nous pouvons vérifier des propriétés de logique de temps linéaire avec des contrantes (CLTL). Dans cette thèse, nous introduisons deux extensions du formalisme de partitions interactives:(1) l'une pour gérer le traitement audio en utilisant le langage de programmation français Faustet (2) l'autre pour traiter des condition et des branchements, permettant de spécifier des choix et des boucles. Pour la première extension, nous présentons une sémantique basée sur les structures d'événements temporisés et des idées sur la façon de définir une sémantique opérationnelle. Pour la deuxième extension, nous présentons une mise en oeuvre et la comparaison des résultats du jitter relative moyenne d'une implémentation d'un arpège base sur l'algorithme de Karplus-Strong par rapport aux implémentations existants écrits dans Pure Data. Nous définissons aussi un format de sauvegarde XML pour les partitions interactives et pour la extension avec branchement conditionnel. Un format de sauvegarde est crucial pour assurer la persistance des partitions. / Technology has shaped the way on which we compose and produce music. Notably, the invention of microphones and computers pushed the development of new music styles in the 20th century. In fact, several artistic domains have been benefiting from such technology developments ; for instance, Experimental music, non-linear multimedia, Electroacoustic music, and interactive multimedia. In this dissertation, we focus on interactive multimedia.Interactive multimedia deals with the design of scenarios where multimedia content and interactive events are handled by computer programs. Examples of such scenarios are multimedia art installations, interactive museum exhibitions, some Electroacoustic music pieces, and some Experimental music pieces. Unfortunately, most interactive multimedia scenarios are based on informal specifications, thus it is not possible to formally verify properties of such systems. We advocate the need of a general and formal model. Interactive scores is a formalism to describe interactive multimedia scenarios. We propose new semantics for interactive scores based on timed eventstructures. With such a semantics, we can specify properties for the system, in particular, properties about traces, which are difficult to specify as constraints. In fact, constraints are an important part of the semantic model of interactive scores because the formalism is based on temporal constraints among the objects of the scenario. We also present an operational semantics of interactive scores based on the non-deterministic timed concurrent constraint (ntcc) calculus and we relate such a semantics to the timed event structures semantics. With the operational semantics, we formally describe the behavior of a score whose temporal object durations can be arbitrary integer intervals. The operational semantics is obtained from the timed event structures semantics of the score. To provide such a translation, we first define the normal form of a timed event structure in which events related with zero-duration delays are collapsed into a single one. We also define the notion of dispatchable timed event structures. Event structures such that its constraint graph can be dispatched by relying only on local propagation.We believe that operational semantics in ntcc offers some advantages over existing Petri nets semantics for interactive scores; for instance, the duration of the temporal objects can be arbitrary integer intervals, whereas inprevious models of interactive scores, such durations can only be intervals to represent equalities and inequalities. In this dissertation, we also introduce two extensions of the formalism of interactive scores : (1) one to handle audio processing using the Fast AUdio Stream (Faust) languageand (2) another one to handle conditional branching, allowing designers to specify choices and loops. For the first extension, we present a timed event structures semantics and ideas on how to define operational semantics. For the second extension, we present an implementation and results comparing the average relative jitter of an implementation ofan arpeggio based on Karplus-Strong with respect to existing implementations of Karplus written in Pure Data. We also define a XML file format for interactive scores and for the conditional branching extension. A file format is crucial to assure the persistence of the scores. Ntcc models of interactive scores are executed using Ntccrt, a real-time capable interpreter for ntcc. They can also be verified automatically using ntccMC, a bounded-time automata based model checker for ntcc which we introduce in this dissertation. Using ntccMC, we can verify properties expressed on constraint linear-time logic. Ntcc has been used in the past, not only for multimedia interaction models, but alsofor system biology, security protocols and robots.

Doctor of Philosophy / Department of Animal Sciences and Industry / Jeffrey S. Stevenson / Four experiments were conducted in beef cows to determine factors that increased the probability of pregnancy per AI when cows are inseminated by appointment. Cows in all experiments were inseminated after a 7-d CO-Synch + CIDR program (100 μg GnRH [2 mL Factrel, Pfizer Animal Health, Whitehouse Station, NJ] 7 d before 25 mg PGF₂[subscript]α [d 0; 5 mL Lutalyse; Pfizer Animal Health]). Experiment 1 compared 1 vs. 2 inseminations and GnRH injection times at 60 and 75 h after the CO-Synch + CIDR program. Delaying AI until 75 h, according to interpretation of estrus-detection patches, for cows not in estrus by 60 h after CIDR insert removal increased (P < 0.05) pregnancy risk (PR) compared with cows not in estrus and inseminated at 60 h (51.4 vs. 41.7%), respectively. The necessity of GnRH injection concurrent with AI was tested in experiment 2. Cows displaying estrus by 65 h that were injected with GnRH had similar PR to cows in estrus and not treated with GnRH (61.9 vs. 60.4%), respectively. Cows in experiment 2 that did not display estrus, but were treated with a GnRH injection at 65 h and then inseminated at 84 h after CIDR insert removal had increased PR compared with similar cows not treated with GnRH (33.4 vs. 15.0%; P < 0.01), respectively. Experiments 3 and 4 were observational studies conducted to determine if blood metabolites glucose and beta-hydroxy butyrate (BHB experiment 3), or physical body and blood metabolites, (glucose, BHB, non-esterified fatty acids [NEFA], blood urea nitrogen [BUN], body weight, rump fat [RF], or BCS; experiment 4) were indicative of future reproductive success in suckled beef cows enrolled in a timed AI program. In experiment 3, plasma glucose concentration 10 d before AI was lesser (P = 0.01; 52.2 vs. 56.9 mg/dL) and serum BHB concentration was lesser (P < 0.01) in cows that became pregnant 35 d after timed AI than for cows that did not become pregnant (600 vs. 690 μM), respectively. Experiment 4 identified relationships between indicators and reproductive success including the finding that serum NEFA concentration 2 to 4 wk before AI is negatively correlated (P < 0.05) with PR to AI.

Nous présentons dans ce travail de thèse des approches pour la spécification et la vérificationdes systèmes temporisés. La première partie concerne une méthode de spécification enutilisant les diagrammes états-transitions pour modéliser un système donné en partant d’unedescription textuelle. Cette méthode guide l’utilisateur pour le développement de la modélisation.Elle comporte plusieurs étapes et utilise des observateurs d’états et des événements afind’engendrer le diagramme états-transitions. Un outil qui implémente les différentes étapes de laméthode de spécification pour une application semi-automatique est présenté. La seconde partieconcerne une traduction des diagrammes états-transitions vers les réseaux de Petri colorés, cequi permet d’utiliser les méthodes de vérification. Nous prenons en considération dans cette traductionun ensemble important des éléments syntaxiques des diagrammes états-transitions, telsque la concurrence, la hiérarchie, etc. Un outil qui implémente la traduction pour un passageautomatique des diagrammes états-transitions vers les réseaux de Petri colorés est en cours de développement.La dernière partie concerne l’intégration des contraintes temporelles dans les deuxapproches précédentes. Nous définissons des annotations pour les diagrammes états-transitionsdont nous fournissons la syntaxe et la sémantique. Ces annotations seront ensuite utilisées dansla méthode de spécification et la traduction. Le but est de proposer des annotations faciles àcomprendre et à utiliser avec une syntaxe qui prend en compte des contraintes parmi les plusutilisées. / In order to specify and verify timed systems, we present in this thesis approaches using UMLstate machines and coloured Petri nets. Our first approach is a specification method that takesinto account a textual description of the system and generates the corresponding state machinediagram. This method helps a non-expert user to model a system in a structural way. We presenta tool that implements the specification method. Our second approach is the translation of UMLstate machine diagrams to coloured Petri nets diagrams. In this approach we take into account animportant set of UML state machine elements that allows the modelling of concurrent systems,etc. A tool that implements the approach and allows us to automate the translation is beingdeveloped. Finally, the last approach is the integration of time constraints in our specificationmethod and in our translation. We propose a set of annotations to model time in state machinediagrams, and we define the corresponding syntax and semantics.

Nowadays, real-time systems are ubiquitous in several application domains.Such an emergence led to an increasing need of performance (resources,availability, concurrency, etc.) and initiated a shift from theuse of single processor based hardware platforms, to large setsof interconnected and distributed computing nodes. This trend introduced the birthof a new family of systems that are intrinsically distributed, namelyemph{Networked Embedded Systems}.Such an evolution stems from the growing complexity of real-time softwareembedded on such platforms (e.g. electronic control in avionicsand automotive domains), and the need to integrate formerly isolated systems so thatthey can cooperate, as well as share resources improving thus functionalitiesand reducing costs.Undoubtedly, the design, implementation and verification of such systems areacknowledged to be very hard tasks since theyare prone to different kinds of factors, such as communication delays, CPU(s)speed or even hardware imprecisions, which increases considerably the complexity ofcoordinating parallel activities.In this thesis, we propose a rigorous design flow intended forbuilding distributed real-time applications.We investigate timed automata based models, with formally defined semantics, in orderto study the behavior of a given system with some imposed timing constraints when deployedin a distributed environment. Particularly, we study emph{(i)} the impact of the communicationdelays by introducing a minimum latency between actions executions and the effectivedate at which actions executions have been decided,and emph{(ii)} the effect of hardware imperfections, more precisely clocks imprecisions,on systems execution by breaking the perfect clocks hypothesis, often adopted duringthe modeling phase. Nevertheless, timed automata formalism is intended to describe a highlevel abstraction of the behavior of a given application.Therefore, we use an intermediate representation ofthe initial application that, besides having say{equivalent} behavior, explicitly expressesimplementation mechanisms, and thus reduces the gap between the modeling and the concreteimplementation. Additionally, we contribute in building such systems by emph{(iii)}proposing a knowledge based optimization method that aims to eliminate unnecessarycomputation time or exchange of messages during the execution.We compare the behavior of each proposed model to the initial high level model and study therelationships between both. Then, we identify and formally characterize the potential problemsresulting from these additional constraints. Furthermore, we propose execution strategies thatallow to preserve some desired properties and reach a say{similar} execution scenario,faithful to the original specifications. / Aujourd'hui, les systèmes temps réel sont omniprésents dans plusieurs domaines.Une telle expansion donne lieu à un besoin croissant en terme de performance (ressources,disponibilité, parallélisme, etc.) et a initié par la même occasion une transition del'utilisation de plateformes matérielles à processeur unique, à de grands ensemblesde nœuds de calcul inter-connectés et distribués. Cette tendance a donné la naissanceà une nouvelle famille de systèmes connue sous le nom de emph{Networked Embedded Systems},qui sont intrinsèquement distribués.Une telle évolution provient de la complexité croissante des logiciels temps réelembarqués sur de telles plateformes (par exemple les système de contrôle en avioniqueet dans domaines de l'automobile), ainsi que la nécessité d'intégrer des systèmes autrefoisisolés afin d'accomplir les fonctionnalités requises, améliorant ainsi les performanceset réduisant les coûts.Sans surprise, la conception, l'implémentation et la vérification de ces systèmes sontdes tâches très difficiles car ils sont sujets à différents types de facteurs, tels que lesdélais de communication, la fréquence du CPU ou même les imprécisions matérielles,ce qui augmente considérablement la complexité lorsqu'il s'agit de coordonner les activités parallèles.Dans cette thèse, nous proposons une démarche rigoureuse destinée à la construction d'applicationsdistribuées temps réel.Pour ce faire, nous étudions des modèles basés sur les automates temporisés, dont la sémantiqueest formellement définie, afin d'étudier le comportement d'un système donné avec des contraintes de tempsimposées lorsqu'il est déployé dans un environnement distribué. En particulier, nous étudionsemph{(i)} l'impact des délais de communication en introduisant une latence minimale entreles exécutions d'actions et la date à laquelle elles ont été décidées,et emph{(ii)} l'effet des imperfections matérielles, plus précisément les imprécisionsd'horloges, sur l'exécution des systèmes.Le paradigme des automates temporisés reste néanmoins destiné à décrire une abstractiondu comportement d'une application donnée.Par conséquent, nous utilisons une représentation intermédiaire del'application initiale, qui en plus d'avoir un comportement say{équivalent}, exprimeexplicitement les mécanismes mis en œuvre durant l'implémentation, et donc réduit ainsil'écart entre la modélisation et l'implémentation réelle.De plus, nous contribuons à la construction de tels systèmes en emph{(iii)}proposant une optimisation basée sur la emph{connaissance}, qui a pour but d'éliminer lestemps de calcul inutiles et de réduire les échanges de messages pendant l'exécution.  Nous comparons le comportement de chaque modèle proposé au modèle initial et étudionsles relations entre les deux. Ensuite, nous identifions et caractérisons formellement lesproblèmes potentiels résultants de ces contraintes supplémentaires. Aussi, nous proposonsdes stratégies d'exécution qui permettent de préserver certaines propriétés souhaitéeset d'obtenir des scénarios d'exécution say{similaires}, et fidèles aux spécificationsde départs.

A necessidade de aumento de produção, da redução de custos e do aumento da qualidade de bens de consumo, tem motivado a constante evolução dos sistemas de produção, migrando os tradicionais sistemas de produção para os modernos e complexos sistemas de manufatura, onde a performance depende da eficiência dos equipamentos e do controle do processo. Por outro lado, a eficiência dos equipamentos depende de sua confiabilidade e manutenabilidade. Neste trabalho a análise de performance é avaliada com o uso de Rede de Petri p-t-Temporizada e através de simulações, incluindo a avaliação da confiabilidade do processo pela análise da otimização da saída do sistema, isto é, quantidade de itens produzidos. Nesta abordagem, uma lógica linear foi desenvolvida e validada utilizando-se uma comparação de resultados das classes de estados do algoritmo proposto com a ferramenta de simulação Tina para um modelo de um esquema produtor consumidor. Apresenta-se um estudo de caso na indústria automotiva, consistindo na análise dos problemas reais enfrentados em uma fábrica de carrocerias, com o uso da Rede de Petri p-t-Temporizada. / The necessity of growing in production, with reduction of costs and improvement in the quality of consumption good, has motivated the constant evolution of production systems, transforming traditional production systems into the modern and complex manufacturing systems, where the performance depends on the efficiency of the equipment and process control. On the other hand, the equipment efficiency depends of their reliability and maintainability. In this work it is proposed a performance evaluation and analysis with the use of p-t- Timed Petri Nets using simulations, including process reliability analysis of the system through the throughput optimization, i.e., produced amount of goods. In this approach, a linear logic statement was developed and validated using a comparison of results of classes of states between the Tina simulation environment and the algorithm considered for a model of a producing consuming system. A case study in the automotive industry is presented, consisting of the analysis of the real problems found in a body shop plant, with the use of Timed Petri Net.