The aim of this thesis is to develop a stereo vision system to locate and classify objects moving on a conveyor belt. The vision system determines the locations of the objects with respect to a world coordinate system and class of the objects. In order to estimate the locations of the objects, two cameras placed at different locations are used. Image processing algorithms are employed to extract some features of the objects. These features are fed to stereo matching and classifier algorithms. The results of stereo matching algorithm are combined with the calibration parameters of the cameras to determine the object locations. Pattern classification techniques (Bayes and Nearest Neighbor classifiers) are used to classify the objects. The linear velocity of the objects is determined by using an encoder mounted to the shaft of the motor driving the conveyor belt. A robot can plan a sequence of motion to pick the object from the conveyor belt by using the output of the proposed system.

In this thesis, the walking of a humanoid system is simulated applying an optimal predictive control algorithm. The simulation is built using Matlab and Simulink softwares. Four separate physical models are developed to represent the single support and the double support phases of a full gait cycle. The models are three dimensional and their properties are analogous to the human&rsquo / s. In this connection, the foot models in the double support phases include an additional joint which connects the toe to the foot. The kinematic relationships concerning the physical models are formulated recursively and the dynamic models are obtained using the Newton &ndash / Euler formulation. The computed torque method is utilized at the level of joints. In the double support phase, the redundancy problem is solved by the optimization of the actuating torques. The command accelerations required to control the gait are obtained by applying an optimal predictive control law. The introduced humanoid walker achieves a sustainable gait by tuning the optimization and prediction parameters. The control algorithm manages the tracking of the predefined walking pattern with easily realizable joint accelerations. The simulation is capable of producing all the reaction forces, reaction moments and the values of the other variables. During these computations, a three dimensional view of the humanoid walker is animated simultaneously. As a result of this study, a suitable simulation structure is obtained to test and improve the mechanical systems which perform bipedal locomotion. The modular nature of the simulation structure developed in this study allows testing the performance of alternative control laws as well.

This work focuses on modeling and real-time control of a motion simulator for dynamic testing of a two-axis gyro-stabilized head mirror used in modern tanks. For this purpose, a six-degree-of freedom Stewart Platform which can simulate disturbances on the stabilized head mirror during operation of the tank is employed. Mathematical models of the Stewart Platform are constructed using MATLAB and ADAMS. Control system infrastructure is constructed and real-time control system elements are employed. Controller tuning is achieved by using the developed mathematical models in MATLAB. These parameters are applied in the real-time control system and fine tuning is achieved. Accuracy of the motion simulator is tested by mounting an Inertial Measurement Unit on the Stewart Platform. Further control system strategies are discussed by means of simulation.

In this thesis, an aeroservoelastic test appratus is designed to investigate the flutter phenomena in a low speed wind tunnel environment. Flutter is an aeroelastic instability that may occur at control surfaces of aircrafts and missiles. Aerodynamic, elastic, and inertial forces are involved in flutter. A mathematical model using aeroelastic equations of motion is derived to investigate flutter and is used as a basis to design the test setup. Simulations using this mathematical model are performed and critical flutter velocities and frequencies are found. Stiffness characteristics of the test setup are determined using the results of these simulations. The test setup is a two degrees of freedom system, with motions in pitch and plunge, and is controlled by a servomotor in the pitch degree of freedom. A NACA 0012 airfoil is used as a control surface in the test setup. Using this setup, the flutter phenomena is generated in Ankara Wind Tunnel (AWT) and experiments are conducted to validate the results of the theoretical aeroelastic mathematical model calculations.

Flutter is one of the most important phenomena in which aerodynamic surfaces become unstable in certain flight conditions. Since the 1930&amp / #8217 / s many studies were conducted in the areas of flutter prediction in design stage, research of design methods for flutter prevention, derivation and confirmation of flutter flight envelopes via tests, and in similar subjects for aircraft wings. With the use of controllers in 1960&amp / #8217 / s, studies on the active flutter suppression began. First the classical controllers were used. Then, with the improvement of the controller synthesis methods, optimal controllers and later robust controllers started to be used. However, there are not many studies in the literature about fully movable control surfaces, commonly referred to as fins. Fins are used as missile control surfaces, and they can also be used as a horizontal stabilizer or as a canard in aircraft. In the scope of this thesis, controllers satisfying the performance and flutter suppression requirements of a fin are synthesized and compared. For this purpose, H2, Hinf, and mu controllers are used. A new flutter suppression method is proposed and used. In order to assess the performance of this method, results obtained are compared with the results of another flutter suppression method given in the literature. or the purpose of implementation of the controllers developed, aeroelastic model equations are derived by using the typical section wing model with thin airfoil assumption. The controller synthesis method is tested for aeroelastic models that are veloped for various flow regimes / namely, steady incompressible subsonic, unsteady incompressible subsonic, nsteady compressible subsonic, and unsteady compressible supersonic.

Chalcogenide based Phase Change Materials are currently of great technological interest in the growing field of optoelectronics. Ge₂Sb₂Te₅ (GST) is the most widely studied phase change material, and it has been commercially used in both optical and electronic data storage applications, due to its ability to switch between two different atomic configurations, at high speed and with low power consumption, as well as its high optical and electrical contrast between amorphous and crystalline states. Despite its well-known optical and electrical properties, the operation in combination of optical and electrical domains has not yet been fully investigated. This work studies the operation of GST nano-devices exposed to a combination of optical and electrical stimuli or mixed mode by asking, is it possible to electrically measure an optically induced phase change, or vice versa? If so, how do the optical and electrical responses relate to each other, and is it possible to operate GST with a combination of optical and electrical signals? What are the technical constraints that need to be considered in order to fabricate GST devices that could be operated either optically or electrically? In order to answer these questions, experiments that characterized the optical and electrical responses of GST based nano-devices were performed. It was found that different crystallization mechanisms may have influence in the response, and that the thermal and optical design characteristics of the device play a key role in its operation. Finally a proof of principle, of an opto-electonic memory device that can be read electrically, reset optically and write electrically, is presented. This opens up possibilities for the development of new opto-eloectronic applications such as non-volatile interfaces between future photonics and electronics, high speed optical communication detectors, high speed cameras, artificial retinas and many more.

Um Daten am effektivsten zu nutzen, muss die moderne Ökonometrie ihren Werkzeugkasten an Modellen erweitern und neu denken. Das Feld, in dem diese Transformation am besten beobachtet werden kann, ist die kausale Inferenz. Diese Dissertation verfolgt die Absicht Probleme zu untersuchen, Lösungen zu präsentieren und neue Methoden zu entwickeln Machine Learning zu benutzen, um kausale Parameter zu schätzen. Dafür werden in der Dissertation zuerst verschiedene neuartige Methoden, welche als Ziel haben heterogene Treatment Effekte zu messen, eingeordnet. Im zweiten Schritt werden, basierend auf diesen Methoden, Richtlinien für ihre Anwendung in der Praxis aufgestellt. Der Parameter von Interesse ist der „conditional average treatment effect“ (CATE). Es kann gezeigt werden, dass ein Vergleich mehrerer Methoden gegenüber der Verwendung einer einzelnen Methode vorzuziehen ist. Ein spezieller Fokus liegt dabei auf dem Aufteilen und Gewichten der Stichprobe, um den Verlust in Effizienz wettzumachen. Ein unzulängliches Kontrollieren für die Variation durch verschiedene Teilstichproben führt zu großen Unterschieden in der Präzision der geschätzten Parameter. Wird der CATE durch Bilden von Quantilen in Gruppen unterteilt, führt dies zu robusteren Ergebnissen in Bezug auf die Varianz. Diese Dissertation entwickelt und untersucht nicht nur Methoden für die Schätzung der Heterogenität in Treatment Effekten, sondern auch für das Identifizieren von richtigen Störvariablen. Hierzu schlägt diese Dissertation sowohl die „outcome-adaptive random forest“ Methode vor, welche automatisiert Variablen klassifiziert, als auch „supervised randomization“ für eine kosteneffiziente Selektion der Zielgruppe. Einblicke in wichtige Variablen und solche, welche keine Störung verursachen, ist besonders in der Evaluierung von Politikmaßnahmen aber auch im medizinischen Sektor wichtig, insbesondere dann, wenn kein randomisiertes Experiment möglich ist. / To use data effectively, modern econometricians need to expand and rethink their toolbox. One field where such a transformation has already started is causal inference. This thesis aims to explore further issues, provide solutions, and develop new methods on how machine learning can be used to estimate causal parameters. I categorize novel methods to estimate heterogeneous treatment effects and provide a practitioner’s guide for implementation. The parameter of interest is the conditional average treatment effect (CATE). It can be shown that an ensemble of methods is preferable to relying on one method. A special focus, with respect to the CATE, is set on the comparison of such methods and the role of sample splitting and cross-fitting to restore efficiency. Huge differences in the estimated parameter accuracy can occur if the sampling uncertainty is not correctly accounted for. One feature of the CATE is a coarser representation through quantiles. Estimating groups of the CATE leads to more robust estimates with respect to the sampling uncertainty and the resulting high variance. This thesis not only develops and explores methods to estimate treatment effect heterogeneity but also to identify confounding variables as well as observations that should receive treatment. For these two tasks, this thesis proposes the outcome-adaptive random forest for automatic variable selection, as well as supervised randomization for a cost-efficient selection of the target group. Insights into important variables and those that are not true confounders are very helpful for policy evaluation and in the medical sector when randomized control trials are not possible.

In dieser Arbeit werden verschiedene Sprechtempoparameter in den Sprachen Chinesisch, Deutsch, Englisch, brasilianisches und europäisches Portugiesisch sowie Spanisch zueinander in Bezug gesetzt: Die Maßeinheiten Laute, Silben, Wörter und Morpheme pro Sekunde werden unter Ein- und Ausschluss der Pausenzeit gemessen. Dabei ergeben sich für die germanischen Sprachen niedrige Durchschnittswerte, die romanischen Sprachen liegen im mittleren bis hohen Bereich und das Chinesische weist je nach betrachtetem Parameter mittlere oder hohe Werte auf. Eine Analyse der Pausenstruktur ergibt vergleichbare Pausenanteile, aber einen höheren Anteil interner Pausen in den Sprachen Chinesisch, europäisches Portugiesisch und Spanisch. Weiterhin werden nonverbale Aspekte berücksichtigt und die Größen Gesten, Gesichts¬ausdrücke und Bewegungen des Kopfes und Oberkörpers pro Minute gemessen. Im brasilianischen Portugiesisch sind Gesten und im Chinesischen Bewegungen des Kopfes und Oberkörpers auffallend häufig. Geschlechtsspezifische Unterschiede sind in allen Bereichen gering und nur im europäischen Portugiesisch statistisch signifikant. Bezüglich des wahrgenommenen Sprechtempos zeigt sich in Experimenten und Umfragen, dass die romanischen Sprachen als schnell bezeichnet werden. Chinesische Hörer schätzen außerdem das Sprechtempo anders ein als deutsche, wobei die Silbenkomplexität für beide Gruppen unterschiedliche Rollen spielt. Daneben bestehen zwischen diesen beiden Gruppen signifikante Unterschiede in der Beurteilung von Sachlichkeit und Vertrauenswürdigkeit in Aufnahmen. Für die untersuchte Sprechsituation wird davon ausgegangen, dass in verschiedenen Sprachgemeinschaften die Verwendung unterschiedlicher sprachlicher (verbaler wie nonverbaler) Mittel bevorzugt wird. Das Sprechtempo wird damit als eine von phonetisch-phonologischen Faktoren, aber auch von in gesellschaftlichen Gruppen vorhandenen Tendenzen im Sprachgebrauch beeinflusste Größe verstanden. / This doctoral thesis is a cross-linguistic analysis of speech tempo, comparing Chinese, English, German, Brazilian and European Portuguese, and Spanish. There are several parameters of speech tempo under analysis: sounds, syllables, words, and morphemes per second. The Germanic languages show lower values, whereas Romance languages tend to have medium and high values, and Chinese appears to have different, both medium and high values depending on what element is counted. An analysis of pause structure renders comparable percentages of time spent on silence, yet the proportion of internal pauses is considerably higher in Chinese, European Portuguese and Spanish. Furthermore, nonverbal aspects of speech are considered and data are collected for the measures gestures, facial expressions and movements of the head and torso per minute. Gestures are remarkably more frequent in Brazilian Portuguese, and movements of the head/torso are highly frequent in Chinese. Differences between genders are generally small, only within European Portuguese some statistically significant differences can be observed. Surveys on stereotypes and perception experiments show that the Romance languages are generally perceived as spoken more rapidly. The listeners’ native language appears to be a factor of perceived speech tempo: Chinese listeners’ responses are significantly different from German listeners’ responses. Syllable complexity plays different roles for these two groups of listeners. They also show significant differences when judging objectivity and trustworthyness in recordings. Considering the particular communication situation analyzed here, it is pointed out that in different speech communities, different verbal and nonverbal behavior is preferred in certain contexts. Speech tempo is influenced by phonetic and phonological factors, as well as tendencies towards a certain communicative style that can be observed within a given society.

本研究以天真避險、傳統OLS模型、OLS共整合模型及Bivariate GARCH模型探討SIMEX及OSE所交易的日經225 (Nikkei 225)股價指數期貨對日本及台灣股市風險的避險效果，測試在台灣股價指數期貨尚未推出之際，投資人是否可能採用鄰近國家，如日本的日經股價指數期貨，來規避台灣股市風險。本研究採用每週三週報酬資料，研究期間自1988年9月3日起至1995年12月底止，全部樣本期間共有376筆資料，劃分為兩個子期間，並以第二子期間做樣本外測試，避險期間分為一週、兩週及四週。 　　實證結果發現： 　　(1) SIMEX日經指數期約、OSE日經指數期約、日經股價指數及台灣股價指數的時間數列均非常態分配。經一階差分之後，上述四個時間數列才會為定態數列。日經股價指數期貨與日經股價指數之間有共整合關係，此乃表示現貨與期貨價格之間存在有長期均衡關係，但日經股價指數期貨與台灣股價指數之間並無共整合關係。 　　(2) Bivariate GARCH模式在各研究期間所得到的各參數的估計值，大多顯著，這說明不論在日本或台灣市場，以日經股價指數期貨規避股票市場風險時，期貨與現貨分配會有隨時間而變動的現象。 　　(3) 在日經指數的現貨市場中： 　　1. OLS共整合模型的避險比率較傳統OLS模型為高。使用SIMEX期貨契約避險所需要的避險比率較使用OSE期貨契約為避險工具時為小，而且不論使用SIMEX或OSE期貨契約避險，當避險期間越長，避險比率越大。 　　2. 在樣本內實證中，以OSE期貨契約避險所造成的投資組合變異數較使用SIMEX期約為大，而且投資組合變異數隨避險期間的增長而有下降的趨勢，但在樣本外的期間中，卻無如此的明顯趨勢。 　　3. 除了在日本股市大崩盤之前的實證期間顯示不論是使用SIMEX或OSE期貨契約，Bivariate GARCH模型的避險效果均較好之外，在其他的實證期間中，GARCH模型大約只比天真避險模式效果好，卻比其他模型效果差，而這種情況在使用OSE期貨契約時更為明顯，不過不論使用哪種模型，都能比不避險時減少大部份現貨的風險。 　　4. 從樣本內實證期間發現SIMEX與OSE契約在避險效果上是有差別的，但樣本外實證卻未發現避險效果上有明顯差別。 　　(4)在台灣股價指數的現貨市場下： 　　1. 不論使用SIMEX或OSE期貨契約避險在崩盤前所需要的避險比率均較崩盤後為高，而不論使用SIMEX或OSE期貨契約避險，避險比率均差不多。 　　2. 樣本內或樣本外實證都發現，若使用天真避險模式避險還不如不避險的好。除了在大崩盤後的樣本內實證中，GARCH模式的組合變異比傳統OLS模式為高之外，Bivariate GARCH模式的確優於其他避險模式。但日經指數期約與台灣股價指數所形成的投資組合變異數比在日本市場時高出甚多，且使用OLS或GARCH模式只能略微降低不避險狀態下所造成的變異數。不論是參數的估計值或避險績效都支持日經指數期貨與台灣股價指數間存在有GARCH效果。

Dans cette thèse, on considère deux types de processus longues mémoires : les processus stationnaires et non-stationnaires. Nous nous consacrons à l'étude de leurs propriétés statistiques, les méthodes d'estimation, les méthodes de prévision et les tests statistiques. Les processus longue mémoire stationaires ont été largement étudiés au cours des dernières décennies. Il a été démontré que des processus longue mémoire ont des propriétés d'autosimilarité, qui sont importants pour l'estimation des paramètres. Nous passons en revue les propriétés d'auto-similairité des processus longue mémoire en temps continu et en temps discret. Nous proposons deux propositions montrant que les processus longue mémoire sont asymptotiquement auto-similaires du deuxième ordre, alors que processus courte mémoire ne sont pas asymptotiquement auto-similaires du deuxième ordre. Ensuite, nous étudions l'auto-similairité des processus longue mémoire spécifiques tels que les processus GARMA à k facteurs et les processus GIGARCH à k facteurs. Nous avons également étudié les propriétés d'auto-similarités des modèles heteroscedastiques et des processus avec des sauts. Nous faisons une revue des méthodes d'estimation des paramètres des processus longue mémoire, par méthodes paramétriques (par exemple, l'estimation par maximum de vraisemblance et estimation par pseudo-maximum de vraisemblance) et les méthodes semiparamétriques (par exemple, la méthode de GPH, la méthode de Whittle, la méthode de Robinson). Les comportements de consistance et de normalité asymptotique sont également étudiés pour ces estimateurs. Le test sur l'ordre fractionnaire intégré de la racine unité saisonnière et non-saisonnière des processus longue mémoire stationnaires est très important pour la modélisation des series économiques et financières. Le test de Robinson (1994) est largement utilisé et appliqué aux divers modèles longues mémoires bien connus. A partir de méthode de Monte Carlo, nous étudions et comparons les performances de ce test en utilisant plusieurs tailles d'échantillons. Ce travail est important pour les praticiens qui veulent utiliser le test de Robinson. Dans la pratique, lorsqu'on traite des données financières et économiques, la saisonnalité et la dépendance qui évolvent avec le temps peuvent souvent être observées. Ainsi une sorte de non-stationnarité existe dans les données financières. Afin de prendre en compte ce genre de phénomènes, nous passons en revue les processus non-stationnaires et nous proposons une nouvelle classe de processus stochastiques: les processus de Gegenbauer à k facteurs localement stationnaire. Nous proposons une procédure d'estimation de la fonction de paramètres en utilisant la transformation discrète en paquets d'ondelettes (DWPT). La robustesse de l'algorithme est étudiée par simulations. Nous proposons également des méthodes de prévisions pour cette nouvelle classe de processus non-stationnaire à long mémoire. Nous dennons des applications sur le terme de la correction d'erreurs de l'analyse cointégration fractionnaire de l'index Nikkei Stock Average 225 et nous étudions les prix mondiaux du pétrole brut.