SHAREOK Repository
Not a member yet
49261 research outputs found
Sort by
A study on diffusion probabilistic models for image generation
Get PDFDiffusion probabilistic models have emerged as powerful tools for image generation and synthesis tasks. This research delves into the intricate relationship between the hyperparameters of these models and the underlying hardware, aiming to provide insight into optimizing performance. The study specifically investigates the effects of key hyperparameters: the number of feature parameters, the number of timesteps, the image size, the number of images in the dataset, the learning rate, and the number of epoch iterations. Additionally, the influence of hardware, particularly GPU memory, on the overall performance is examined. Through a systematic experimentation framework, we consider different hyperparameters and hardware configurations to quantify their impact on model convergence, image generation quality, and computational efficiency. The research aims to identify optimal hyperparameter settings for diverse tasks while considering the constraints imposed by available hardware resources. Moreover, the study explores potential trade-offs and synergies between hyperparameter tunning and hardware specifications, shedding light on the interplay between algorithmic choices and computational capabilities. The findings from this research contribute to a nuanced understanding of how diffusion probabilistic models can be fine-tuned for image generation, considering the practical implications of hardware limitations. By bridging the gap between algorithmic design and hardware constraints, this work provides valuable guidance for practitioners seeking to leverage diffusion models effectively in real-world scenarios which require image generation
Ultrawideband antenna solutions for weather radar applications
No full textThe success of future multi-functional communication, radar, sensing, and countermeasure
systems heavily rely on ultrawideband (UWB) antennas. UWB antennas
have a very high demand in electromagnetic compatibility (EMC) testing,
radar calibration, satellite tracking, reflector feeds, remote sensing, wideband instrument
for snow measurements (WISM), medical imaging, through-wall imaging,
and UWB radar systems. In an ideal case, these multi-faceted systems require
the use of a single UWB “one-size-fits-all” antenna that satisfies the bandwidth and
radiation requirements compared to using multiple narrowband antennas saving
time, labor, and overall cost of the system.
This dissertation delves into the theory, design, and feasibility of one of the
kinds of a single UWB antenna or a single integrated UWB antenna system for
their potential use in the characterization or calibration of weather radars or other
UWB communication systems. Firstly, detailed Maxwell-Garnett particle mixing
analytical models are presented and evaluated based on their performance and
error analysis. The correct analytical model is identified and used in the synthesis
of tunable artificial dielectric material (TADM) at microwave (S-band) and mm-Wave
(W-band) frequencies for applications such as RF substrates for antennas
and microwave components such as filters. This technique allows for the utilization
of cost-effective materials that already offer outstanding thermal properties and
seamless compatibility with a multi-layer fabrication process, can now be equipped with desired tunable εeff and loss tangent characteristics. The measured results
show substantial reductions of 45% in εeff and 38% in loss tangent values in the
S-band, along with reductions of 32% and 72% in εeff and tanδ, respectively,
in the mm-Wave frequency band. Impressively, the simulated, calculated, and
measured results demonstrate an excellent level of agreement.
To see the feasibility of using commercially available state-of-the-art UWB
antennas for in-situ characterization of radars, a detailed analysis is carried out
based on the idea of using a single UWB probe antenna for radars operating in
the L-, S-, C-, X-, Ku-, and K-bands. This analysis shows in detail that using
commercially available UWB antennas is not suitable for in-situ measurements
due to their wide beamwidth performance, especially for weather radars operating
in the most commonly used S-band at 3 GHz. To solve this challenge, several
solutions are presented. Among the solutions presented based on using corrugated
horns and multi-layer or hyperbolic dielectric lenses, the Maxwell-Garnett theory
that laid the foundation of the design of tunable artificial dielectrics is also applied
in the synthesis of a UWB graded-index (GRIN) artificial dielectric lens antenna
(ADLA). Proposed novel solutions make the use of these commercially available
antennas viable for radars and communication systems.
A UWB open-boundary dual-polarized quad-ridged horn antenna (QRHA)
having an impressive absolute bandwidth of more than 30 GHz (32:1) covering 1-
32 GHz, is then presented making it an ideal choice for UWB communication and
radar systems covering a wide range of applications relying on common frequency
bands including L, S, C, X, Ku, and K. The unique design allows for a favorable
impedance matching, acceptable stable and narrow-beamwidth performance, and
most importantly no-pattern degradation at higher frequencies which is a common
concern for most ridged-horn antennas designed thus far
Fighting “Firewater:” Native American Temperance Reform and Federal Indian Policy in the Nineteenth Century
No full textIn 1802, Little Turtle, Chief of the Miami, wrote to President Thomas Jefferson explaining the increased presence of alcohol in Indian Country. This sparked a century of Indigenous peoples fighting to control alcohol and assert their sovereignty in the face of growing encroachment from the United States. Natives saw the detrimental impact alcohol had on not only their tribal sovereignty but also on how non-Natives viewed their communities. Restricting the whiskey trade and reforming alcohol policy helped the tribes' ability to resist the non-Native groups that used alcohol for coercion and to perpetuate the “drunken Indian” myth. Throughout the nineteenth and twentieth centuries, American Indian nations and the United States federal government contested the boundaries of sovereignty. Individuals often apply American Indian sovereignty, or the belief in tribal power and authority, to a tribe’s ability to assert influence over their political and legal system. Rather, this dissertation argues that sovereignty during the nineteenth century took on many forms, and the fight for temperance is illustrative of how tribes asserted their sovereignty through cultural and societal processes. By focusing on temperance, this work offers a fresh approach to sovereignty by demonstrating how tribes used their agency to assert cultural sovereignty and self-determination through reforming their society in the face of an expanding United States. This is not a study in Native alcoholism; rather it argues that Indigenous peoples used temperance as a way to assert their sovereignty and strengthen their sense of nationalism and self-determination. The relationship between sovereignty and temperance intersects with other reform movements and shifting Indian policy during the nineteenth century. Indigenous Nations did not rely on the U.S. federal government to regulate the trade of alcohol within Indian Country. They instead took it upon themselves and used their tribal sovereignty to regulate the whiskey trade within their territory
Effect of Curing Temperature on Nanomodified Fiber-Reinforced Polymer Composites
Get PDFDispersion of pristine and functionalized-COOH carbon nanotubes (CNTs) into the resin matrix used to develop fiber-reinforced polymer (FRP) composites is known to improve FRP properties such as shear strength, UV resistance, strength, and stiffness. Nanomodification of the FRP matrix with CNTs is also known to improve bonds within the FRP. However, there is still a gap in knowledge on the effect of temperature on curing characteristics when CNTs are incorporated. In this study, mechanical testing and material characterization of nanomodified FRP composites was conducted to identify the effects of curing under room temperature (30 °C) and elevated temperature (110 °C). The resin for FRP composites was nanomodified with pristine and functionalized multi-walled carbon nanotubes using a standard dispersion protocol. FRP composites were fabricated using vacuum assisted hand layup techniques and prepared for ASTM standard testing. Static tensile testing and interfacial adhesion tests were conducted to evaluate the mechanical performance. Differential scanning calorimetry and thermogravimetric analysis were performed to determine curing characteristics to inform on the polymerization of nanomodified resins cured under the two temperature conditions. Scanning electron microscopy was performed to identify CNT dispersion characteristics. It was found that curing FRP composites with nanomodified resins at elevated temperatures increased the tensile and interfacial adhesion strength and stiffness and also reduced ductility. It can be understood from this study how target performance metrics in a wide range of structural applications can be achieved in FRP composites by incorporating nanomodified resins cured at varying temperatures
Open Access Datasets from Federal Government Agencies
Get PDFThe U.S. Government is the largest publisher in the world and produces primary source data and statistics. Federal information is open and free and without copyright. Data and data sets published by U.S. government agencies are open access and are available to researchers to use. Learn the key access points to finding federal data.N
Developing a Framework for Seamless Prediction of Subseasonal to Seasonal Extreme Precipitation Events in the United States
No full textExtreme precipitation events are natural hazards that pose a significant risk to life and a commensurately large cost through property loss across multiple timescales. Unfortunately, extreme precipitation, especially extended-duration precipitation, remains as one of the most challenging hazards to forecast. The subseasonal to seasonal timescale, defined as the period between 2 weeks and 3 months, is characterized with low skill and a subsequent lack of informative forecast products for end-users. The present work constitutes a major step into the identification, characterization, prediction, and communication of subseasonal extreme precipitation. To identify an extreme precipitation period, thresholds for both total precipitation and the duration of the precipitation are used to identify events with sufficient length to accentuate the synoptic and subseasonal contributions. I then generate databases of 14-day extreme precipitation periods over the contiguous United States (CONUS) using observational datasets, reanalysis products, and climate model simulations, displaying the flexibility of my developed scheme. These datasets are used to quantify trends and synoptic-scale characteristics in subseasonal extreme precipitation. At present, the observational database is fully accessible in tabular format along with informational fact sheets describing the definition and climatology all available online. Rossby wave trains are shown to be present and co-located with anomalously strong moisture in several regions during both observed and simulated extreme periods. The observational database is also used to fit random forests and convolutional neural networks to understand the predictability of these extreme periods. The random forests show some skill above climatology in differentiating extreme and non-extreme days in three different regions of the U.S. with better predictability over the West Coast compared to the Central Great Plains. Although the critical success index (CSI) scores were relatively low, peaking at 0.23 along the West Coast, the random forests appear more skillful when using skill scores tailored towards extreme events. The convolutional neural networks produce probabilities of subseasonal extreme precipitation onto a map over the CONUS are verified using an object-oriented scheme (e.g., centroid offset, area ratio, etc.) and CSI scores reach 0.3. Although the skill is seemingly low, the networks are often identifying "close hits" and may still bring value to end users who are more tolerant to false alarms. The object-oriented scheme allows for flexibility in matching forecasts and observations that end-users can tolerate and gives a better indication of forecast usability. Although much work lies ahead to develop an operational forecast product, the work herein lays the foundation for future works to continually make advances into the S2S timescale
Causal failures and cost-effective edge augmentation in networks
Get PDFNode failures have a terrible effect on the connectivity of the network. In traditional models, the failures of nodes affect their neighbors and may further trigger the failures of their neighbors, and so on. However, it is also possible that node failures would
indirectly cause the failure of nodes that are not adjacent to the failed one. In a power grid, generators share the load. Failure of one generator induces extra load on other generators in the network, which could further trigger their failures. We call such
failures causal failures. In this dissertation, we consider the impact of causal failures on multiple aspects of one network. More specifically, we list the content as follows.
• In Chapter 1, we introduce basic concepts of networks and graphs, classical models of failures and formally define causal failures in a given network.
• Chapter 2 addresses the network’s robustness and aims to find the maximum number of causal failures while maintaining a connected component with a size of at least a given integer. More specifically, we are looking into the number of causal node failures we can tolerate yet have most of the system connected with α being used to parametrize.
• Chapter 3 deals with vulnerability, wherein we aim to find the minimum number of causal failures such that there are at least k connected components remaining. We are looking for the set of causal failures that will result in the network being disconnected into k or more components.
• In Chapter 4, we consider causal node failures occurring in a cascading manner. Cascading causal node failures affect communication within nodes, which is dependent on the paths that connect them. Therefore, in this context of the cascading causal failure model, we study the impact of cascading causal failures on the distance between a pair of nodes in the network. More precisely, given a network G, a set of causal failures (containing possible cascading failures), a pair of nodes s and t, and a constant α ≥ 1, we would like to determine the maximum number of causal failures that can be applied (meaning that the nodes in the causal failures are removed), such that in the resulting network G′, dG′ (s, t) ≤ α × dG(s, t), where dG(s, t) and dG′ (s, t) are the distance between nodes s and t in the networks G and G′, respectively.
• In Chapter 5, we consider causal edge failures in flow networks and investigate the impact of causal edge failures on flow transmission. We formulate an optimization problem to find the maximum number of causal edge failures after which the flow network can still deliver d units from source node s to terminal node t.
• In Chapter 6, we consider edge-weighted network augmentation when facing causal failures. We look for a set of edges with minimum weight such that the network maintains an α-giant component when applying each causality individually. We show that the optimization problems in these chapters are NP-hard and provide the corresponding mixed integer linear programming models. Moreover, we design polynomial-time heuristic algorithms to solve them approximately. In each chapter, we run experiments on multiple synthetic and real networks to compare the performance of the mixed integer linear programming models and the heuristic algorithms. The results show that the heuristic algorithms show their efficacy and efficiency compared
to the mixed-integer linear programming models
Big Brother in Blue: The Prying Eyes of the Police Surveillance State and Perceptions Shaped by Violent Crime Victimization
Get PDFPolice surveillance technologies have become increasingly prevalent in contemporary Western society, raising public concerns about privacy and the illegitimate police use of such surveillance. Still, there is little existing research concerning individual perceptions of the institution of mass police surveillance (i.e., the police surveillance state), and even less exploring the impact of violent crime victimization and police legitimacy on support for/opposition toward the police surveillance state, respectively. Using national public survey data, I employ Structural Equation Modeling (SEM) to examine how crime victimization and other sociodemographic characteristics influence perceptions of the police surveillance state, and to investigate if these relationships are predicated upon the mediative impact of police legitimacy. I find that experiencing violent crime victimization negatively shapes attitudes toward the police surveillance state overall, as compared to those having never experienced such victimization; however, when violent crime victimization influences understandings of the police as more legitimate, they subsequently support the police surveillance state at an increasing rate as compared to non-victims. This study highlights the significance of victimization experiences amidst a growing body of literature on public perceptions of the police and policing surveillance. The present research thus implicates the importance of amplifying diverse and vulnerable voices in discussions of equitable policing practices, and ultimately shaping governmental/departmental policies and regulating efficacious police-community relations
Flow structure effects on turbulent transport using computations
No full textTurbulent flow is the most typical flow in various fields such as biological, chemical and industrial applications and atmospheric phenomena. The main goal of this research is to investigate the fundamental effects of the very large-scale structures of turbulence on convective transport and the effects of the interplay between molecular diffusion and flow structure on turbulent transport. Applying Lagrangian computations in conjunction with direct numerical simulations for turbulent flow in channel and Couette flow, it could explore the effects of the very large-scale turbulent fluid motions on scalar transport and determine the mechanism by which they contribute to transport. From that we can develop a predictive model for turbulent scalar transfer and control this.
One of our applications is the effect of turbulent flow in cardiovascular devices such as blood pumps, ventricular assist devices (VADs) and even heart valves. According to clinical findings, VAD patients often suffer from acquired von Willebrand disease after VAD implantation. Indeed, the von Willebrand Factor protein is susceptible to supraphysiological shear stress and extensional stress which leads to the conformational change of vWF. Apart from that, the detailed measurement of stresses in a hemodynamic velocity field is difficult, especially in-situ for turbulent flow conditions. In such cases, the use of computational methods has become critical for both the probing of the hemodynamic conditions and for device design. Therefore, recognizing advantages of Direct Numerical Simulation (DNS) and Lagrangian Scalar Tracking (LST) we applied it to the case of vascular assist devices and indicate (i) the development of a numerical methodology for obtaining the history of the stresses on vWF molecules as they move in a flow field; (ii) the detailed calculation of shear and extensional stress statistical distributions on the vWF molecules showcasing the importance of the tails of the distributions in addition to average values; and (iii) the investigation of the importance of the flow field configuration and of the location of vWF injection on the distribution of stresses over time. This study’s contribution is the statistical data for hydrodynamic stress on vWF along the trajectories of protein particles that could contribute to design the medical devices and reduce the cost of experiments.
The second application is to investigate the role of helicity in turbulent transport in channel and Couette flow and the interplay between helicity with scalar transport and coherent structures. Helicity is also important for understanding the relationship between coherent structures and turbulent kinetic energy. There has been a hypothesis that coherent structures at different scales are associated with regions in which helicity is large and dissipation of turbulent kinetic energy is low. Low dissipation means that these coherent structures could survive for a long time leading to significant impacts on the flow. In this study, Direct Numerical Simulation (DNS) of turbulent Poiseuille and Couette flow were used in combination with Lagrangian Scalar Tracking (LST) at various Schmidt numbers of 0.7, 6 and infinite (i.e., fluid particles), and the friction Reynolds number for both simulations was 300 to probe the correlation between helicity and dissipation, and the role of helicity in the regions of coherent structures in anisotropic turbulence. Our Lagrangian Scalar Tracking could provide the fluid velocity at each marker location with time. From these data, one can compute turbulent kinetic energy dissipation, vorticity, and helicity along these trajectories. The autocorrelation coefficients, the cross-correlation coefficients and the joint probability density function are employed to investigate the relation between helicity and dissipation, and helicity and vorticity and with the vertical velocity in the Lagrangian framework. In addition, conditional statistics for scalar markers are evaluated in flow regions of high dissipation or coherent structure regions based on vortex identification criteria. In addition, scalar markers that dispersed most or least in the flow field were also calculated to provide more evidence for the role of helicity in transport in turbulent flow
Examining the Vertical Structure of Hurricane Laura (2020) Using Azimuthal-Mean Vertical Profiles and an EOF Analysis
Get PDFThe eyewall, inner rainband, and outer rainband regions of tropical cyclones (TCs) have been extensively studied, resulting in numerous conceptual models. While these conceptual models are insightful regarding basic-state kinematics in TCs, they do not detail kinematic variability. Therefore, kinematic perturbations within TCs are not well documented. One of the University of Oklahoma Shared Mobile Atmospheric Research and Teaching Radars (SRs) was deployed near the Louisiana coast at the time Hurricane Laura made landfall on the Louisiana coast early on 27 August 2020. The SR and the Lake Charles, LA WSR-88D (KLCH) sampled the northeast quadrant of Laura in their collective dual-Doppler lobe from 0000 UTC until KLCH sustained substantial damage at around 0550 UTC. With a 1-km x 1-km x 0.5-km spatial resolution and a temporal resolution of 5-10 minutes, the dual-Doppler analyses (DDAs) captured the eyewall, inner rainband, and outer rainband regions of the hurricane. This dataset therefore provides an opportunity to study the kinematic variability of all three TC regions. This research serves to validate the conceptual models of previous literature using DDA-mean azimuthal-mean vertical profiles. It also aims to uncover the vertical structure of kinematic variability across the TC using an empirical orthogonal function (EOF) analysis. Additionally, this EOF analysis quantifies the amount of variance explained by each perturbation profile. The results of this research largely agree with the conceptual models, though some features are not supported and others are found to not represent the majority of the variance