abstract: A critical problem for airborne, ship board, and land based radars operating in maritime or littoral environments is the detection, identification and tracking of targets against backscattering caused by the roughness of the sea surface. Statistical models, such as the compound K-distribution (CKD), were shown to accurately describe two separate structures of the sea clutter intensity fluctuations. The first structure is the texture that is associated with long sea waves and exhibits long temporal decorrelation period. The second structure is the speckle that accounts for reflections from multiple scatters and exhibits a short temporal decorrelation period from pulse to pulse. Existing methods for estimating the CKD model parameters do not include the thermal noise power, which is critical for real sea clutter processing. Estimation methods that include the noise power are either computationally intensive or require very large data records. This work proposes two new approaches for accurately estimating all three CKD model parameters, including noise power. The first method integrates, in an iterative fashion, the noise power estimation, using one-dimensional nonlinear curve fitting, with the estimation of the shape and scale parameters, using closed-form solutions in terms of the CKD intensity moments. The second method is similar to the first except it replaces integer-based intensity moments with fractional moments which have been shown to achieve more accurate estimates of the shape parameter. These new methods can be implemented in real time without requiring large data records. They can also achieve accurate estimation performance as demonstrated with simulated and real sea clutter observation datasets. The work also investigates the numerically computed Cram\'er-Rao lower bound (CRLB) of the variance of the shape parameter estimate using intensity observations in thermal noise with unknown power. Using the CRLB, the asymptotic estimation performance behavior of the new estimators is studied and compared to that of other estimators. / Dissertation/Thesis / Doctoral Dissertation Electrical Engineering 2019

This work develops and examines design strategies for reducing the stiffness of 3D-printed compliant mechanisms. The three aspects of a flexure that determine its stiffness are well known: material, boundary conditions, and geometry. In a highly constrained design space however, flexure stiffness may remain unacceptably high even while arriving at the limits of design constraints. In this work, changes to geometry and boundary conditions are examined that lead to drastically reduced stiffness behavior without changing flexure thickness, width, or length. Changes to geometry can result in very complex mechanisms. However, 3D printing enables almost arbitrarily complex geometries. This dissertation presents three design strategies for stiffness reduction: static balancing, lattice flexures, and compound joints. Static balancing refers to changes in the boundary conditions that result in a near-zero net change in potential energy storage over the useful deflection of a flexure. In this work, I present a method for static balancing that utilizes non-dimensional parameters to quickly synthesize a joint design with stiffness reduced by nearly 90%. This method is not only simple and straightforward, it is applicable to a wide range of flexure topologies. The only requirements on the joint to be balanced are that it must be approximated as a pin joint and torsion spring, and it must have a well-understood stiffness when subjected to a compressive load. Lattice flexures result from modifications to geometry that reduce cross-sectional area without changing width or thickness. However, the reduction in stiffness is greater than the reduction in cross sectional area. This can occur because the bending load is now carried by beams partially in torsion. Two lattice geometries are proposed and analyzed in detail using analytic and numeric techniques. It is shown that the off-axis stiffness behavior of lattice flexures can be better than that of conventional blade flexures while bending stiffness is reduced >60%. Compound joints are those that consist of arrays of flexures arranged co-axially. This arrangement provides increased range of motion, generally decreased stiffness, and improved stability. Additionally, a method is herein presented to reduce the parasitic center shift of a compound joint to nearly zero at a specified deflection. The penultimate chapter demonstrates how all three strategies can be used together, and includes new results to facilitate their combination.

Ma recherche consiste en une analyse sémantique et morphologique des formes verbales dites « surcomposées » en français (j’ai eu fait, j’avais eu fait, j’aurai eu fait, etc.). La forme la plus fréquemment employée et la mieux représentée est la forme appelée « passé surcomposé », du type j’ai eu fait. Mais à des degrés divers, toutes les formes du paradigme sont attestées.Le passé surcomposé connaît deux emplois distincts. Le premier, appelé « standard », est attesté dans toute la francophonie. Cette forme peut être décrite comme un passé résultatif, car, par son aspect « accompli », elle pointe la phase qui, dans le passé, fait suite au procès. Utilisée en relation avec le passé composé à valeur aoristique, qui pointe le procès en cours de déroulement, elle crée des effets d’antériorité (quand il a eu mangé, il est parti).Le second emploi, dit « régional », n’est attesté que dans les domaines à substrat occitan et francoprovençal. Ce second passé surcomposé peut être décrit comme un parfait d’expérience, car il possède toujours une valeur « expérientielle », qui indique qu’il est arrivé au moins une fois à une situation donnée de se produire à l’intérieur d’un certain intervalle temporel. Pour les locuteurs qui emploient ces formes, un énoncé comme j’ai eu mangé du requin signifie ainsi il m’est arrivé au moins une fois de manger du requin. Les formes « standard » et les formes « régionales » ne sont pas seulement distinctes sémantiquement. Elles sont également différentes sur le plan morphologique : tandis que les premières sont construites par composition de l’auxiliaire (avoir eu + fait), les secondes se construisent par l’insertion du morphème eu, marqueur expérientiel, dans le syntagme verbal composé : avoir (+eu) fait. Cette différence n’est certes pas « visible » pour les formes qui se construisent avec l’auxiliaire « avoir ». Mais elle se révèle avec les verbes qui demandent l’auxiliaire « être » : les formes résultatives se construisent en effet sur le modèle de avoir été + parti (quand il a été parti, je suis allée me coucher), tandis que les formes expérientielles se construisent sur le modèle de être (+eu) parti (on est eu partis en vacances en Italie). La correspondance entre sens et forme étant absolue, la thèse défendue dans mon travail est qu’il ne s’agit pas seulement de types d’emploi distincts, mais bien de formes verbales différentes / My research consists in a semantic and morphological analysis of the so-called “surcomposé” verbal forms in French (j’ai eu fait, j’avais eu fait, j’aurai eu fait, etc.). The most frequently used and best-represented form is called “passé surcomposé”, such as j’ai eu fait. However, all the forms of the paradigm are attested to some degree.The passé surcomposé displays two distinct uses. The first one, known as the “standard use”, is attested throughout the French-speaking world. This type can be described as a resultative past. Indeed, through its “perfect” aspect, it points to the phase, in the past, following the process. Used in association with the passé composé with an aoristic value, which points to the on-going process, it creates anteriority effects (quand il a eu mangé, il est parti).The second use, referred to as “regional”, is only attested in the areas with an Occitan and a Franco-Provencal substrate. This second passé surcomposé can be described as an experiential perfect because it always has an “experiential” value that indicates that a given situation has occurred at least once within a certain temporal interval. Speakers who use these forms understand an utterance like j’ai eu mangé du requin as I have eaten shark at least once.The “standard” and “regional” forms do not only differ in terms of semantics. They are also different at the morphological level: whereas the former are constructed by auxiliary composition (avoir eu + fait), the latter are constructed by the insertion of the morpheme eu, which is an experiential marker, in the composed verbal phrase: avoir (+eu) fait. Of course this difference is not “visible” for forms constructed with the auxiliary “avoir”. It is however revealed with verbs requiring the auxiliary “être”: resultative forms are constructed on the model avoir été + parti (quand il a été parti, je suis allée me coucher), while experiential forms are constructed on the model être (+eu) parti (on est eu partis en vacances en Italie).As the form-meaning correspondence is absolute, the present work proposes a view according to which these forms correspond to different verbal forms and not only distinct uses

Information contained in chemical compounds, fingerprint databases, social networks, and interactions between websites all have one thing in common: they can be represented as graphs. The need to analyze, compare, and classify graph datasets has become more evident over the last decade. The graph isomorphism problem is known to belong to the NP class, and the subgraph isomorphism problem is known to be an NP-complete problem. Several error-tolerant graph matching techniques have been developed during the last two decades in order to overcome the computational complexity associated with these problems. Some of these techniques rely upon similarity measures based on the topology of the graphs. Random walks and edit distance kernels are examples of such methods. In conjunction with learning algorithms like back-propagation neural networks, k-nearest neighbor, and support vector machines (SVM), these methods provide a way of classifying graphs based on a training set of labeled instances. This thesis presents a novel approach to error-tolerant graph matching based on current flow analysis. Analysis of current flow in electrical networks is a technique that uses the voltages and currents obtained through nodal analysis of a graph representing an electrical circuit. Current flow analysis in electrical networks shares some interesting connections with the number of random walks along the graph. We propose an algorithm to calculate a similarity measure between two graphs based on the current flows along geodesics of the same degree. This similarity measure can be applied over large graph datasets, allowing these datasets to be compared in a reasonable amount of time. This thesis investigates the classification potential of several data mining algorithms based on the information extracted from a graph dataset and represented as current flow vectors. We describe our operational prototype and evaluate its effectiveness on the NCI-HIV dataset.

Abstract Multiphase Droplet Interactions with a Single Fiber By: Noor M. Farhan A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy at Virginia Commonwealth University. Virginia Commonwealth University, 2019 Director: Hooman V. Tafreshi, Professor, Department of Mechanical and Nuclear Engineering Formulating the physics of droplet adhesion to a fiber is interesting intellectually and important industrially. A typical example of a droplet–fiber system in nature is the dew droplets on spider webs, where droplets first precipitate and grow on the fibers, but they eventually fall when they become too heavy. Obviously, quantifying the force of adhesion between a droplet and a fiber is crucial in designing fog harvesting devices or manufacturing filtration media for liquid–gas or liquid–liquid separation, among many other industrial applications. This study is aimed at developing a mathematical framework for the mechanical forces between a droplet and a fiber in terms of their physical and wetting properties. To this end, a series of experiments were conducted to detach ferrofluid droplets of varying volumes from fibers with different diameters and Young–Laplace contact angles (YLCAs) in a controlled magnetic field. The force of detachment was measured using a sensitive scale and used along with the results of numerical simulations to develop a semi-analytical expression for the force required to detach a droplet from a fiber. This universally-applicable expression allows one to predict the force detachment without the need to run an experiment or a computer simulation. This work also reports on the use of magnetic force to measure the force of detachment for nonmagnetic droplets for the first time. This is accomplished by adding a small amount of a ferrofluid to the original nonmagnetic droplet to create a compound droplet with the ferrofluid nesting inside or cloaking the nonmagnetic droplet. The ferrofluid is then used to induce a body force to the resulting compound droplet and thereby detach it from the fiber. The recorded detachment force is used directly (the case of nesting ferrofluid) or after scaling (the case of cloaking ferrofluid) to obtain the force of detachment for the original nonmagnetic droplet. The accuracy of these measurements was examined through comparison with numerical simulations as well as available experimental data in the literature. In addition, a simple method is developed to directly measure the intrinsic contact angle of a fiber (i.e., Young–Laplace Contact angle of the fiber material) with any arbitrary liquid. It is shown that the intrinsic contact angle of a fiber can be obtained by simply measuring the angle between the tangent to the fiber surface and the tangent to the droplet at the contact line, if the droplet possesses a clamshell conformation and is viewed from the longitudinal direction. The novelty of the proposed method is that its predictions are not affected by the volume of the droplet used for the experiment, the wettability of the fiber, the surface tension of the liquid, or the magnitude of the body force acting on the droplet during the experiment. Also, a liquid droplet interaction with granular coatings is simulated and the droplet apparent contact angle (ACA) and the transition from Cassie (fully dry) to Wenzel (fully wet) state as a function to the roughness wavelength have been studied. For a fixed droplet volume, two different granular coatings have been used, spherical and hemispherical bumps. It is demonstrated that the chemistry (YLCA) and geometrical parameters for the granular microtexture play an important effect on the droplet ACA and its transition from Cassie to Wenzel state.

No description available.

Baiting with sodium monofluoroacetate (1080) to protect lambs (Ovis aries) from red foxes (Vulpes vulpes L.) has become more frequent in NSW and other parts of Australia during the last 10 years despite the lack of reported evidence evaluating the effects of fox baiting on lamb survival. NSW Agriculture has developed fox control recommendations aimed at minimising impact, but these guidelines have not been tested experimentally. Defining the extent of a pest problem and the effectiveness of pest control are key components of a strategic approach to vertebrate pest management as it is the damage of pests that justifies their control. This thesis describes an experimental evaluation of the recommended practice of fox control in NSW. The effects of three levels of fox control were tested in the experiment; no treatment, baiting once a year before lambing (the recommended practice) and baiting three times a year (thought to be the maximum farmers would instigate). Each treatment had two replicates. No previous manipulative experiment using synchronous controls and matched replicates has been undertaken to test the effects. The study quantifies the level of fox predation on healthy lambs and the level of predation on lambs that had other causes including illness and mismothering contributing to this fox predation. It also examines the response of the fox population, lamb predation and lambing outcomes to different levels of fox control. The cost effectiveness of fox control is examined in relation to lamb predation and an investigation of the optimum level of fox control is begun. The experiment also provides the first chance to consider the examination of multiple response variables and the scale of field ecology experiments required to recognise a significant response and avoid a Type II error due to between replicate variability even with tightly controlled site selection criteria to standardise experimental sites, and with the synchrony of experimental control and treatment surveys. The study occurred on five sheep properties near Boorowa (34°28'S, 148°32'E) and Murringo (34°18'S, 148°3 1'E) in south-eastern Australia. The terrain was undulating to hilly with a maximum elevation of 660 m above sea level. The main agricultural enterprises in the district are Merino wool, fat lamb and beef cattle production and winter cereal cropping. The native vegetation of Eucalyptus woodland has been mostly cleared, though remnant patches occur. Most of the area is now sown with pasture of Phalaris tuberosa, Lolium spp. and clover Trifolium spp.. The experimental properties grazed self-replacing Merino flocks, primarily for wool production, so lamb survival was vital to the economic operation of the farm. Over 50 selection criteria including lamb survival rates, ewe fertility and bloodline, sheep management practices, climate and habitat features that affect lamb survival, past fox control practices and prey species were used to select sites Sites were representative of most sheep farming properties in the region, but were also extremely similar in factors that affected fox abundance and ewe and lamb survival, thus minimising variation between replicate sites. The manufactured meat baits used to poison foxes contained 3 mg of sodium monofluoroacetate (compound 1080). A replacement baiting program was carried out in 1995 and 1996. Fox control programs were carried out over the experimental units and adjacent buffer zones covering approximately two fox territories, approximately 6km2, around the lambing paddock under study. The recommended fox control practices described by NSW Agriculture also included neighbouring farmers taking part in an extended group baiting program. In all the area baited at varying intensities totalled 3400 km2. Synchronised lambing with neighbours was a further recommended practice to reduce fox predation and was carried out on these sites. Lambing occurred during a six to eight week period in late winter on all sites, a practice known as 'spring lambing', and on many surrounding properties so a surplus of lambs was available to foxes over a relatively short time. The benefits of fox control were measured directly as enhanced lamb survival derived from differences in lamb marking rates between ultrasounded flocks of approximately 1000 ewes and the predation of lambs was measured from over 2000 lamb carcases post-mortemed in 1994, 1995 and 1996. A mean of 138 lambs were expected at ultrasounding from 100 ewes and 113 lambs per 100 ewes were alive at lamb marking. There was no significant (P>0.05) effect of fox control on lambing performance (the number of lambs per 100 ewes that lambed) Fox predation was inferred as the cause of lamb death in a minimum of 0.8% and a maximum of 5.3% of lamb carcases during 1995 and 1996. There was a significant (P<0 05) effect of fox control on the minimum possible percentage of lamb carcases classified as healthy lambs killed by foxes, with the percentage declining from 1.50% (no fox control), to 0.90% (fox control once per year) to 0.25% (fox control three times per year). There was also a significant (P<0.005) effect of fox control on the maximum possible percentage of lamb carcases classified as healthy lambs killed by foxes with the percentage declining from 10.25% (no fox control), to 6.50% (fox control once per year) to 3.75% (fox control three times per year). The observed results were used to estimate the number of treatment replicates needed to be confident of detecting an effect of predator control on lamb marking performance. The estimated numbers were very high if small effects were to be detected. No significant correlation between the fox density and the minimum and maximum possible number of lambs carcases classified as killed by foxes was found. Bait uptake was monitored as were the costs of fox control.

長久以來，漢語中詞彙的並列關係已經受到許多漢語學者的注意。本文以大約可被認為是秦代之前成書的十本史書及子書作為研究資料，對當時詞彙的並列使用作一些研究。本文共分為六章。第一章是動機、範圍、材料及內容簡述。第二章對於「並列結構」一詞的界定作一鑑別標準及說明，並將以前學者的研究意見作整理。另外，也簡單說明並列結構的基本詞數是兩個單音詞。第三章討論並列結構的語法功用，發現並列結構在動詞組合和名詞組合之間的語法功用是不同的。而趨向動詞並列結構也可由其特徵分作五類。第四章說明詞序的情形。並列結構中有百分之十有詞序顛倒的情形。第五章研究聲韻對詞序的影響。我們發現，除了聲調以外，聲母和韻母也對詞序有影響。第六章是結論。

Thesis (M.S.)--Rochester Institute of Technology, 2007. / Typescript. Includes bibliographical references.

<p>Over the last years, the emission and on board diagnostics legislations for heavy duty trucks are getting more and more strict. An accurate engine model that is possible to execute in the engine control system enables both better diagnosis and lowered emissions by better control strategies. </p><p>The objective of this thesis is to extend an existing mean value diesel engine model, to include turbo compound. The model should be physical, accurate, modular and it should be possible to execute in real time. The calibration procedure should be systematic, with some degree of automatization. </p><p>Four different turbo compound models have been evaluated and two models were selected for further evaluation by integration with the existing model. The extended model showed to be quite insensitive to small errors in the compound turbine speed and hence, the small difference in accuracy of the tested models did not affect the other output signals significantly. The extended models had better accuracy and could be executed with longer step length than the existing model, despite that more complexity were added to the model. For example, the mean error of the intake manifold pressure at mixed driving was approximately 3.0%, compared to 5.8% for the existing model. The reasons for the improvements are probably the good performance of the added submodels and the systematic and partly automatized calibration procedure including optimization.</p>