Theses and Dissertations from UMD
Permanent URI for this communityhttp://hdl.handle.net/1903/2
New submissions to the thesis/dissertation collections are added automatically as they are received from the Graduate School. Currently, the Graduate School deposits all theses and dissertations from a given semester after the official graduation date. This means that there may be up to a 4 month delay in the appearance of a give thesis/dissertation in DRUM
More information is available at Theses and Dissertations at University of Maryland Libraries.
Browse
14 results
Search Results
Item TOWARDS AUTONOMOUS VERTICAL LANDING ON SHIP-DECKS USING COMPUTER VISION(2022) Shastry, Abhishek; Datta, Anubhav; Chopra, Inderjit; Aerospace Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)The objective of this dissertation is to develop and demonstrate autonomous ship-board landing with computer vision. The problem is hard primarily due to the unpredictable stochastic nature of deck motion. The work involves a fundamental understanding of how vision works, what are needed to implement it, how it interacts with aircraft controls, the necessary and sufficient hardware, and software, how it differs from human vision, its limits, and finally the avenues of growth in the context of aircraft landing. The ship-deck motion dataset is provided by the U.S. Navy. This data is analyzed to gain fundamental understanding and is then used to replicate stochastic deck motion in a laboratory setting on a six degrees of freedom motion platform, also called Stewart platform. The method uses a shaping filter derived from the dataset to excite the platform. An autonomous quadrotor UAV aircraft is designed and fabricated for experimental testing of vision-based landing methods. The entire structure, avionics architecture, and flight controls for the aircraft are completely developed in-house. This provides the flexibility and fundamental understanding needed for this research. A fiducial-based vision system is first designed for detection and tracking of ship-deck. This is then utilized to design a tracking controller with the best possible bandwidth to track the deck with minimum error. Systematic experiments are conducted with static, sinusoidal, and stochastic motions to quantify the tracking performance. A feature-based vision system is designed next. Simple experiments are used to quantitatively and qualitatively evaluate the superior robustness of feature-based vision under various degraded visual conditions. This includes: (1) partial occlusion, (2) illumination variation, (3) glare, and (4) water distortion. The weight and power penalty for using feature-based vision are also determined. The results show that it is possible to autonomously land on ship-deck using computer vision alone. An autonomous aircraft can be constructed with only an IMU and a Visual Odometry software running on stereo camera. The aircraft then only needs a monocular, global shutter, high frame rate camera as an extra sensor to detect ship-deck and estimate its relative position. The relative velocity however needs to be derived using Kalman filter on the position signal. For the filter, knowledge of disturbance/motion spectrum is not needed, but a white noise disturbance model is sufficient. For control, a minimum bandwidth of 0.15 Hz is required. For vision, a fiducial is not needed. A feature-rich landing area is all that is required. The limits of the algorithm are set by occlusion(80\% tolerable), illumination (20,000 lux-0.01 lux), angle of landing (up to 45 degrees), 2D nature of features, and motion blur. Future research should extend the capability to 3D features and use of event-based cameras. Feature-based vision is more versatile and human-like than fiducial-based, but at the cost of 20 times higher computing power which is increasingly possible with modern processors. The goal is not an imitation of nature but derive inspiration from it and overcome its limitations. The feature-based landing opens a window towards emulating the best of human training and cognition, without its burden of latency, fatigue, and divided attention.Item Nonlinear and Stochastic Analysis of Miniature Optoelectronic Oscillators based on Whispering-Gallery Mode Modulators(2021) Nguewou-Hyousse, Helene; Chembo, Yanne K.; Electrical Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Optoelectronic oscillators are nonlinear closed-loop systems that convert optical energy into electrical energy. We investigate the nonlinear dynamics of miniature optoelectronic oscillators (OEOs) based on whispering-gallery mode resonators. In these systems, the whispering-gallery mode resonator features a quadratic nonlinearity and operates as an electrooptical modulator, thereby eliminating the need for an integrated Mach-Zehnder modulator. The narrow optical resonances eliminate as well the need for both an optical fiber delay line and an electric bandpass filter in the optoelectronic feedback loop. The architecture of miniature OEOs therefore appears as significantly simpler than the one of their traditional counterparts, and permits to achieve competitive metrics in terms of size, weight, and power (SWAP). Our theoretical approach is based on the closed-loop coupling between the optical intracavity modes and the microwave signal generated via the photodetection of the output electrooptical comb. In the first part of our investigation, we use a slowly-varying envelope approach to propose a time-domain model to analyze the dynamical behavior of miniature OEOs. This model takes into account the interactions among the intracavity modes, as well as the coupled interactions with the radiofrequency (RF) microstrip. The stability analysis allows us to determine analytically and optimize the critical value of the feedback gain needed to trigger self-sustained oscillations. It also allows us to understand how key parameters of the system such as cavity detuning or coupling efficiency influence the onset of the radiofrequency oscillation. Furthermore, we determine the threshold laser power needed to trigger oscillations in amplifierless miniature OEOs based on WGM modulators. This latter architecture, while also improving on the size, weight, performance and cost (SWAP-C) constraints, is intended to reduce noise in the system. In the second part of our investigation, we use a Langevin approach to perform a stochastic analysis of our miniature OEO. We propose a stochastic mathematical model to describe the system dynamics and analyze the stochastic behavior below threshold. We also propose a normal form approach for the noise power density and the phase noise spectrum. Our study is complemented by time-domain simulations for the microwave and optical signals, which are in excellent agreement with the analytical predictions. In the third part of our study, we discuss our preliminary results in the analysis of the effects of dispersion in a microcomb oscillator with optical feedback. For this purpose, we propose a closed-loop miniature optical oscillator. The output signal is optically amplified before being coupled back into the cavity using a prism coupling. Using a Lugiato-Lefever approach, we propose a spatiotemporal nonlinear partial differential equation to describe the dynamics of the total intracavity field. We perform temporal and spatial analysis and derive the bifurcation maps in anomalous and normal dispersion regimes.Item Designing for the Human in the Loop: Transparency and Control in Interactive Machine Learning(2020) Renner, Alison Marie; Boyd-Graber, Jordan; Findlater, Leah; Computer Science; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Interactive machine learning techniques inject domain expertise to improve or adapt models. Prior research has focused on adapting underlying algorithms and optimizing system performance, which comes at the expense of user experience. This dissertation advances our understanding of how to design for human-machine collaboration--improving both user experience and system performance--through four studies of end users' experience, perceptions, and behaviors with interactive machine learning systems. In particular, we focus on two critical aspects of interactive machine learning: how systems explain themselves to users (transparency) and how users provide feedback or guide systems (control). We first explored how explanations shape users' experience of a simple text classifier with or without the ability to provide feedback to it. Users were frustrated when given explanations without means for feedback and expected model improvement over time even in the absence of feedback. To explore transparency and control in the context of more complex models and subjective tasks, we chose an unsupervised machine learning case, topic modeling. First, we developed a novel topic visualization technique and compared it against common topic representations (e.g., word lists) for interpretability. While users quickly understood topics with simple word lists, our visualization exposed phrases that other representations obscured. Next, we developed a novel, ``human-centered'' interactive topic modeling system supporting users' desired control mechanisms. A formative user study with this system identified two aspects of control exposed by transparency: adherence, or whether models incorporate user feedback as expected, and stability, or whether other unexpected model updates occur. Finally, we further studied adherence and stability by comparing user experience across three interactive topic modeling approaches. These approaches incorporate input differently, resulting in varied adherence, stability, and update speeds. Participants disliked slow updates most, followed by lack of adherence. Instability was polarizing: some participants liked it when it surfaced interesting information, while others did not. Across modeling approaches, participants differed only in whether they noticed adherence. This dissertation contributes to our understanding of how end users comprehend and interact with machine learning models and provides guidelines for designing systems for the ``human in the loop.''Item Adjunct Control: Syntax and processing(2018) Green, Jeffrey Jack; Williams, Alexander; Linguistics; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)This dissertation analyzes the syntax and processing of adjunct control. Adjunct control is the referential relation between the implicit (PRO) subject of a non-finite adjunct clause and its understood antecedent, as in the temporal adjunct in ‘Holly1 went to bed [after PRO1 drinking milk]’, or the rationale clause in ‘August1 sat on the couch [in order PRO1 to read library books]’. Adjunct control is often assumed to involve a syntactic ‘Obligatory Control’ (OC) dependency, but I show that some adjuncts also permit what is referred to as ‘Non-Obligatory Control’ (NOC), as in the sentences ‘The food tasted better [after PRO drinking milk]’ and ‘The book was checked out from the library [in order PRO to read it]’, where PRO refers to some unnamed entity. I argue that for some adjuncts, OC and NOC are not in complementary distribution, contrary to assumptions of much prior literature, but in agreement with Landau (2017). Contrary to implicit assumptions of Landau, however, I also show that this OC/NOC duality does not extend to all adjuncts. I outline assumptions that Landau’s theory would have to make in order to accommodate the wider distribution of OC and NOC in adjuncts, but argue that this is better accomplished within the Movement Theory of Control (Hornstein, 1999) by relaxing the assumption that all adjuncts are phases. Even in adjuncts where both OC and NOC are possible, OC is often strongly preferred. I argue that this is in large part due to interpretive biases in processing. As a foundational step in examining what these processing biases are, the second part of this dissertation uses visual-world eyetracking to compare the timecourse of interpretation of subject-controlled PRO and overt pronouns in temporal adjuncts. The results suggest that PRO can be interpreted just as quickly as overt pronouns once the relevant bottom-up input is received. These experiments also provide evidence that structural predictions can facilitate reference resolution independent of next-mention predictions.Item The Control of a Mathematical Analog of a Tentacle(2015) Xu, Xin; Levine, William S; Electrical Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)A 2-dimensional dynamic analog of squid tentacles was presented. The tentacle analog consists of a multi-cell structure, which can be easily replicated to a large scale. Each cell of the model is a quadrilateral with unit masses at the corners. Each side of the quadrilateral is a spring-damper system in parallel. The spring constants are the controls for the system. The dynamics are subject to the constraint that the area of each quadrilateral must remain constant. The system dynamics was analyzed, and various equilibrium points were found with different controls. Then these equilibrium points were further determined experimentally, demonstrated to be asymptotically stable. A simulation built in MATLAB was used to find the convergence rates under different controls and damping coefficients. Finally, a control scheme was developed and used to drive the system to several configurations observed in real tentacle.Item Development of a Quadcopter Test Environment and Research Platform(2015) De Prins, Christian; Martins, Nuno C; Electrical Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)This thesis first uses a model-based systems engineering approach to model, design, and implement a quadcopter test environment and research platform (TERP). TERP provides quadcopter state information, using a motion capture system, which can be used with custom feedback strategies to enable controlled flight. Next, it makes use of control theory to develop two controllers for quadcopter flight trajectory tracking: one based on linear quadratic regulation (LQR) and one based on model reference adaption. Simulations of both controllers are done in MATLAB using Simulink and seek to demonstrate the improved performance of the adaptive controller over the LQR controller in flight trajectory tracking with payload uncertainties. Flight tests with the LQR controller are then done to validate the TERP System.Item Flight Dynamics Simulation Modeling and Control of a Large Flexible Tiltrotor Aircraft(2014) Juhasz, Ondrej; Celi, Roberto; Aerospace Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)A high order rotorcraft mathematical model is developed and validated against the XV-15 and a Large Civil Tiltrotor (LCTR) concept. The mathematical model is generic and allows for any rotorcraft configuration, from single main rotor helicopters to coaxial and tiltrotor aircraft. Rigid-body and inflow states, as well as flexible wing and blade states are used in the analysis. The separate modeling of each rotorcraft component allows for structural flexibility to be included, which is important when modeling large aircraft where structural modes affect the flight dynamics frequency ranges of interest, generally 1 to 20 rad/sec. Details of the formulation of the mathematical model are given, including derivations of structural, aerodynamic, and inertial loads. The linking of the components of the aircraft is developed using an approach similar to multibody analyses by exploiting a tree topology, but without equations of constraints. Assessments of the effects of wing flexibility are given. Flexibility effects are evaluated by looking at the nature of the couplings between rigid-body modes and wing structural modes and vice versa. The effects of various different forms of structural feedback on aircraft dynamics are analyzed. A proportional-integral feedback on the structural acceleration is deemed to be most effective at both improving the damping and reducing the overall excitation of a structural mode. A model following control architecture is then implemented on full order flexible LCTR models. For this aircraft, the four lowest frequency structural modes are below 20 rad/sec, and are thus needed for control law development and analysis. The impact of structural feedback on both Attitude-Command, Attitude-Hold (ACAH) and Translational Rate Command (TRC) response types are investigated. A rigid aircraft model has optimistic performance characteristics, and a control system designed for a rigid aircraft could potentially destabilize a flexible one. The various control systems are flown in a fixed-base simulator. Pilot inputs and aircraft performance are recorded and analyzed.Item Resource-Based Spike Mitigation Stochastic Control Problems(2013) McCready, Keith; Blankenship, Gilmer; Electrical Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Resource-based spike mitigation stochastic control problems are a class of stochastic control problems, akin to inventory control problems or linear quadratic regulator (LQR) problems. These problems involve consuming a resource to mitigate large, fast losses to a primary state ("spikes"). These properties, included with stochastic elements, draw out a unique behavior where optimal control policies conserve resources during "lucky streaks" and spend the resources during "unlucky streaks." However, these problems often have too many time steps and states to compute an optimal control policy with dynamic programming. This thesis gives examples of such problems and demonstrates how to effectively approximate solutions using suboptimal control methods.Item Recalculating Adjunct Control(2011) ODED, ILKNUR; Lasnik, Howard; Linguistics; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)This study investigates properties of adjunct control with a particular focus on Turkish providing an analysis for different types of adjunct control structures such as temporal adjunct clauses and purpose clauses, which have been understudied in Turkish linguistics. In analyzing adjunct control structures, I use Agree-based Theory of Control (ATC) (Landau 2000 and 2004) as a theoretical basis. I introduce a new interarboreal operation that I call Interarboreal Agree which draws upon the intuitions of Nunes (1995) that syntactic relations can be established between two unconnected trees. This analysis refines ATC in that ATC in its current form fails to account for Obligatory Control reading in adjunct control structures. An important overarching theme of this dissertation is the role of Aspect in determining control properties of adjunct clauses. As an example, I account for the structures that I call SOC (Subject or Object Control) structures in Turkish temporal adjunct clauses by assuming that these clauses do not have an Aspect Phrase projection. I also argue that Case variation in languages that have morphologically-dependent secondary predicates, that is to say, secondary predicates that agree with the NP they predicate in Case, gender or number, can be explained by the presence or absence of an Aspect Phrase projection. Aspect properties of adjunct clauses come into play in purpose clauses as well. For instance, in English, control in purpose clauses exhibits optionality in terms of the choice of the controller, which is not the case in the Turkish counterpart of the same type of purpose clauses. I argue that this is due to the fact that English purpose clauses do not have an Aspect Phrase projection.Item Dynamics and Control of Non-smooth Systems with Applications to Supercavitating Vehicles(2011) Nguyen, Vincent Phuc; Balachandran, Balakumar; Mechanical Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)The subject matter of this dissertation relates to the dynamics of non-smooth vehicle systems, and in particular, supercavitating vehicles. These high-speed underwater vehicles are designed to have sustained vaporous or ventilated gas cavities that form over the entire vehicle. In terms of the modeling, the system non-smoothness is caused by the interaction forces generated when the vehicle contacts the cavity. These planing interactions can cause stable and unstable dynamics, some of which could be limit-cycle dynamics. Here, planing forces are considered on the basis of non-cylindrical cavity shapes that include shifts induced by the cavitator angle of attack. Incorporating these realistic physical effects into a vehicle system model generates a unique hydrodynamic non-smoothness that is characterized by non-constant switching boundaries and non-constant switched dynamics. Nonlinear stability analyses are carried out, Hopf bifurcations of equilibrium solutions are identified, and stabilizing control is investigated. Also considered is partially cavitating system dynamics, where active fin forces are used to support the vehicle. Non-steady planing is also considered, which accounts for vehicle motions into the cavity, and this planing provides a damping-like component in the planing force formulation. Modeled with non-steady planing is a physical time delay relating to the fact that the cavity, where planing occurs, is based on the previous cavitator position and orientation data. This delay is found to be stabilizing for certain values of speed. Maneuvering is considered by using inner-loop and outer-loop control schemes. A feedback inner-loop scheme helps reject fast planing instabilities, while a numeric optimal control approach is used to generate outer-loop commands to guide the vehicle through desired maneuvers. The maneuvers are considered for operations with tight body to cavity clearance, and in which planing is prevalent. Simple search algorithms along with a penalty method for handling the constraints are found to work the best due to the complexity of the non-smooth system dynamics.