Search

The main focus of this thesis lies on image-based rendering (IBR) techniques designed to operate in real-world environments and special attention is paid to the state-of-the-art end-to-end pipelines used to create and display virtual reality (VR) of 360° real-world environments. Head-mounted displays (HMDs) enable users to experience virtual environments freely, but the creation of real-world VR experiences remains a challenging interdisciplinary research problem. VR experiences can greatly differ depending on the used underlying scene representation and the meaning of real-world VR heavily depends on the context, i.e., the system or format at hand. Terminology and fundamental concepts are introduced which are needed to understand related IBR and learned IBR (neural) approaches, which are categorically surveyed in the context of end-to-end pipelines to create real-world IBR experiences. The applicability of the discussed approaches to create real-world VR applications is categorised into practical aspects covering capture, reconstruction, representation, and rendering, which yields a fairly good overview of the research landscape to which this thesis contributes. The life cycle of immersive media production depends on computer vision and computer graphics problems and describes, in its whole, end-to-end pipelines for creating 3D photography used to render high-quality real-world VR experiences. Vision is needed to obtain viewpoint and scene information to create scene representations, i.e., 3D photographs, and computer graphics is needed for creating high-quality novel viewpoints, for instance by applying IBR techniques to the reconstructed scene representation. Lack of widely available immersive real-world VR content which suits current generations of HMDs motivates research in casual 3D photography. Furthermore, augmenting widely available real-world VR formats, e.g., omnidirectional stereo (ODS), seems intriguing in order to increase the immersion of currently available real-world VR experiences. This thesis contributes three end-to-end IBR pipelines for the creation and display of immersive 360° VR experiences, all outperforming the current de-facto standard (ODS) while only relying on moderate computational resources which are commonly available to casual consumers, and one learned IBR approach based on conditional adversarial nets that takes a casually captured video sweep as input to perform high-quality video extrapolation. The ability to casually capture 3D photography might have a profound impact on the way consumers capture, edit, share, and re-live personal experiences in a near foreseeable future.

Electrical power grids are undergoing a significant transformation as part of widespread grid modernization efforts. The integration of renewable energy sources and increased demand-side flexibility are propelling the prevalence of inverter-based resources (IBRs) and distributed energy resources (DERs) across the grid. While these advancements foster a more sustainable and efficient power infrastructure, they also introduce new operational challenges. Traditional model-based methods struggle to cope with the increasing uncertainty and variability inherent in modern power systems. These challenges necessitate the development of new models and algorithms that are not only faster and more reliable, but also scalable to accommodate the expanding network. In order to cope these challenges, we advocate that the widespread installation of sensors such as phasor measurement units (PMUs), alongside advances in computation and communication technologies, have set the stage for a paradigm shift. Specifically, we consider new approaches that harnesses data-driven and machine learning (ML) methodologies to provide efficient, reliable, and scalable solutions for power system operations. The first part of this dissertation focuses on the power system dynamic study. Specifically, we advocate ambient data analytics approach for recovering dynamic responses. Our fully data-driven algorithm can effectively recover pairwise system impulse responses by incorporating all types of ambient data. Note that our method is scalable and does not require any system parameters, while requiring only minimal PMU deployment. Using the recovered impulse responses, we develop an efficient data-driven forced oscillation (FO) localization framework with flexibility in both measurement types and locations. To tackle the modeling challenges from the growing prevalence of IBRs in power system dynamic analysis, we propose a unified approach for learning the dynamics of both synchronous generators (SGs) and IBRs as individual dynamic components. The second part of this dissertation focuses on the graph learning for static optimization problems. For steady state operations, we advocate a topology-aware graph neural network (GNN) based framework for ac-OPF problems, which is faster than direct solvers. Comparing with previous learning methods, our GNN model is more scalable, and systematically tackled the feasibility and topology adaptivity for the first time. For voltage optimization in distribution grids, our scalable and risk-aware learning model can be trained efficiently by focusing on the samples with higher conditional value-at-risk (CVaR) risk, and can predict safe and near-optimal solutions even under communication constraints.

To facilitate experimentation in the field of gas turbine disk dynamics, a new integrally bladed disk, or blisk, was designed and manufactured for use at The Ohio State University (OSU) Gas Turbine Laboratory. The axisymmetric disk portion of the equipment and the blades are manufactured as a single unit in blisks. This new blisk is referred to hereafter as the OSU Rotor 67 Blisk. The OSU Rotor 67 Blisk was designed with traditional airworthy gas turbine rotating hardware standards in mind, although it is not meant for flight. The standards put forth by the Federal Aviation Administration (FAA) are loosely followed to produce a blisk that is similar to hardware that could be used in general aviation. The guidelines put forth by the FAA for rotor design are found in 14 CFR Part 33, and provide requirements for safety aspects of rotor design. The airfoil design and mounting configuration were strongly influenced by the existing NASA Rotor 67 airfoil and disk geometry, with modifications to convert it from a bladed disk to a blisk. Additionally, damper grooves were added to the design to facilitate a variety of damping experiments. After the OSU Rotor 67 Blisk was machined, balanced, and inspected, it was received at OSU, where it was subsequently tested on a non-rotating bench setup to characterize its properties. A roving modal hammer was used along with a calibrated capacitance probe for dynamics measurements. The data generated through the dynamics experiments provided clarity on the natural frequencies and the mistuning of the blisk. These data will be used to design future experiments at the OSU Gas Turbine Laboratory where the blisk will be rotated at different operating speeds and excited by air-jets to evaluate different damping technologies.

An improved spatial statistical approach and probabilistic prediction method for mistuned integrally bladed rotors is proposed and validated with a large population of rotors. Prior work utilized blade-alone principal component analysis to model spatial variation arising from geometric deviations contributing to forced response mistuning amplification. Often, these studies considered a single rotor measured by contact probe coordinate measurement machines to assess the predictive capabilities of spatial statistics through principal component analysis. The validity of the approach has not yet been demonstrated on a large population of mistuned rotors representative of operating fleets, a shortcoming addressed in this work. Furthermore, this work improves the existing predictions by applying principal component methods to sets of airfoil (rotor) measurements, thus effectively capturing blade-to-blade spatial correlations. In conjunction with bootstrap sampling, the method is validated with a set of 40 rotors and quantifies the subset size needed to characterize the population. The work combines a novel statistical representation of rotor geometric mistuning with that of probabilistic techniques to predict the known distribution of forced response amplitudes.

High cycle fatigue of jet aircraft engine turbomachinery components is a serious problem for aircraft engine designers. State-of-the-art compressor designs have reduced mechanical damping and increased unsteady aerodynamic interaction between blade rows. New manufacturing techniques have enabled the rotor to be comprised of an integrally bladed disk, called a blisk. Advanced compressor designs have blades with low aspect ratios, highly complex geometry and higher tip speeds. Inherent in the single disk stage designs is a reduction in mechanical damping; no longer do the connections of blade to disk offer a viable source of damping. Thus, it is important to accurately portray the structural response of the component in a laboratory environment. In an engine environment, a blade row is simultaneously excited from a number of sources. This multiple mode environment was simulated by the modification of the Traveling Wave test rig at the Turbine Engine Fatigue Facility of the Air Force Research Laboratory. The objective of the research presented herein is to measure response of the ADLARF fan rotor excited in a multiple mode environment. The ADLARF fan is significantly structurally mistuned through the replacement of two opposing metallic blades with composite blades done in a past experiment. In this substantially mistuned environment, the assumption of the linear superposition theory for these multiple mode excitations was validated through spatial discrete Fast Fourier transforms and a data point superposition. It has been shown that the response of a bladed disk depends on the engine order of the excitation. Showing the necessity for matching the spatial mode order of the response with that of the engine order of the excitation has provided valuable insight towards fatiguing a complete bladed disk in a laboratory setting.