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

Search Results

Now showing 1 - 10 of 21

Affective Reactions to Uncertainty as Driven by Past Experiences, Personality, and Perceived Valence
(2022) Ellenberg, Molly Deborah; Kruglanski, Arie; Psychology; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
The assumption that uncertainty is inherently threatening which underlies decades of research belies the fact that people rarely react negatively to uncertain situations about which they do not care, and that some are excited by uncertainty. I propose that affective reactions to uncertainty are driven not by uncertainty itself, but by people’s expectations of positive and negative outcomes to personally relevant uncertain situations. I find that positive past experiences predict higher optimism and higher resilience, both of which predict higher tolerance of uncertainty and more positive perceptions of uncertain events. I also find that negative past experiences predict higher pessimism and lower resilience, both of which predict higher intolerance of uncertainty and more negative perceptions of uncertain events. The second study suggests that optimistic people are more likely to approach, rather than avoid, uncertainty. The third study finds that mindfulness training, which emphasizes non-attachment to outcomes, results in more neutral reactions to uncertainty. Theoretical and practical implications are discussed.
Single- and Multi-Objective Feasibility Robust Optimization under Interval Uncertainty with Surrogate Modeling
(2022) Kania, Randall Joseph; Azarm, Shapour; Mechanical Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
This dissertation presents new methods for solving single- and multi-objective optimization problems when there are uncertain parameter values. The uncertainty in these problems is considered to come from sources with no known or assumed probability distribution, bounded only by an interval. The goal is to obtain a single solution (for single-objective optimization problems) or multiple solutions (for multi-objective optimization problems) that are optimal and “feasibly robust”. A feasibly robust solution is one that remains feasible for all values of uncertain parameters within the uncertainty interval. Obtaining such a solution can become computationally costly and require many function calls. To reduce the computational cost, the presented methods use surrogate modeling to approximate the functions in the optimization problem.This dissertation aims at addressing several key research questions. The first Research Question (RQ1) is: How can the computational cost for solving single-objective robust optimization problems be reduced with surrogate modelling when compared to previous work? RQ2 is: How can the computational cost of solving bi-objective robust optimization problems be improved by using surrogates in concert with a Bayesian optimization technique when compared to previous work? And RQ3 is: How can surrogate modeling be leveraged to make multi-objective robust optimization computationally less expensive when compared to previous work? In addressing RQ1, a new single-objective robust optimization method has been developed with improvements over an existing method from the literature. This method uses a deterministic, local solver, paired with a surrogate modelling technique for finding worst-case scenario of parameter configurations. Using this single-objective robust optimization method, improved large-scale performance and robust feasibility were demonstrated. The second method presented solves bi-objective robust optimization problems under interval uncertainty by introducing a relaxation technique to facilitate combining iterative robust optimization and Bayesian optimization techniques. This method showed improved feasibility robustness and performance at larger problem sizes over existing methods. The third method presented in this dissertation extends the current literature by considering multiple (beyond two) competing objectives for surrogate robust optimization. Increasing the number of objectives adds more dimensions and complexity to the search for solutions and can greatly increase the computational costs. In the third method, the robust optimization strategy from the bi-objective second method was combined with a new Monte Carlo approximated method. The key contributions in this dissertation are 1) a new single-objective robust optimization method combining a local optimization solver and surrogate modelling for robustness, 2) a bi-objective robust optimization method that employs iterative Bayesian optimization technique in tandem with iterative robust optimization techniques, and 3) a new acquisition function for robust optimization in problems of more than two objectives.
DATA-DRIVEN PREDICTION, DESIGN, AND CONTROL OF SYSTEM BEHAVIOR USING STATISTICAL LEARNING
(2021) Zhao, Xiangxue; Azarm, Shapour; Balachandran, Balakumar; Mechanical Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
The goal in this dissertation is to develop new data-driven techniques for prediction, design, and control of the behavior of a variety of engineering systems. The data used can be obtained from a variety of sources, including from offline, high-fidelity system’s simulation, physical experiments, and online, sparse measurements from sensors. Three inter-related research directions are followed in this dissertation. Following the first direction, the author presents a multi-step-ahead prediction technique for evaluating a single-response (or single-output of the) system’s behavior through an integration of the data obtained offline from the system’s high-fidelity simulation, and online from single sensor measurements. With regard to the first research direction, the key contribution includes a reasonably fast and accurate prediction strategy that can be used, among others, for online, multi-step ahead forecasting of the system’s operational behavior. Building on the work from the first direction, the author follows a second research direction to present a multi-step ahead prediction technique, this time for a multi-response system’s behavior, that can be used for evaluating various system’s designs and corresponding operations. Data in this case is obtained from the offline, high-fidelity system’s simulations, and online sparse measurements from multiple sensors (or limited number of physical experiments). The main contribution for this second direction is in construction of a new data-driven, multi-response prediction framework that has a robust predictive capability. Along the third research direction, a data-driven technique is used for prediction and co-optimization of a system’s design and control. The data in this case is obtained from sensor measurements or a simulator. The main contribution achieved through the third direction is a new data-driven reinforcement learning-based prediction and co-optimization approach. The methods from this dissertation have numerous applications, including those demonstrated here: (i) assessment of safe aircraft flight conditions (Chapters 2 and 3), (ii) evaluation of design and operation of a robotic appendage (Chapter 3), and (iii) design and control of a traffic system (Chapter 4).
Bayesian Methodology for Reliability Growth Planning and Projection for Discrete-Use Systems Utilizing Multi-Source Data
(2021) Nation, Paul John; Modarres, Mohammad; Mechanical Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
This research aims to present a Bayesian model for reliability growth projection and planning for discrete-use systems suitable for use throughout all stages of system development. Traditional discrete-use models for reliability growth utilize test data from individual test events at the current stage of development. They often neglect the inclusion of historical information from previous tests, testing similar systems, or eliciting expert opinion. Examining and using data attained from prior bench analyses, sub-system tests, or user trial events often fails to occur or is conducted poorly. Additionally, no current approach permits the probabilistic treatment of the initial system reliability at the commencement of the test program in conjunction with the management variables that may change throughout the execution of the test plan.This research aims to contribute to the literature in several ways. Firstly, a new Bayesian model is developed from first principles, which considers the uncertainty surrounding discrete-use systems under arbitrary corrective action regimes to address failure modes. This differs from current models that fail to address the randomized times that corrective actions to observed failure modes may be implemented depending on the selected management strategy. Some current models only utilize the first observed failure on test, meaning a significant loss of information transpires if subsequent failures are ignored. Additionally, the proposed strategy permits a probabilistic assessment of the test program, accounting for uncertainty in several management variables. The second contribution seeks to extend the Bayesian discrete-use system projection model by considering aspects of developmental, acceptance, and operational testing to formulate a holistic reliability growth plan framework that extends over the entire system lifecycle. The proposed approach considers the posterior distribution from each phase of reliability growth testing as the prior for the following growth test event. The same methodology is then employed using the posterior from the final phase of reliability growth testing as the prior for acceptance testing. It then follows that the acceptance testing posterior distribution forms the prior for subsequent operational testing through a Bayesian learning method. The approach reduces unrealistic and unattainable reliability demonstration testing that may result from a purely statistical analysis. The proposed methodology also permits planning for combined developmental and acceptance test activities within a financially constrained context. Finally, the research seeks to define an approach to effectively communicate developmental system reliability growth plans and risks to decision makers. Like many of their other specialist science peers, reliability professionals are fantastic communicators – with other reliability practitioners. However, when reliability professionals move beyond their world to make an impact, they often face the same challenge scientists from every discipline face – the difficulties of clearly communicating science to their audience.
Sampling Based Motion PLanning for Minimizing Position Uncertainty with Stewart Platforms
(2021) Ernandis, Ryan; Otte, Michael; Aerospace Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
The work described in this dissertation provides a unique approach to error based motion planning. Originally designed specifically for use on a parallel robot,these methods can be extended to a more general case of any well-defined robotic platforms. Requirements for application of these methods are a known method of kinematics for defining the system as well as a means of calculating noise based on the system. Two methods of error tracking and two motion planning algorithms are tested here as approaches to this problem. Shown within are the results of the motion planning methods used. One combination of motion planning algorithm and error tracking works best as a general solution to this problem and is designed to work on a parallel robot; specifically, a Stewart platform. The motivation for use of a Stewart platform comes from research done at NASA Langley Research Center in the field of In-Space Assembly.
On Engineering Risks Modeling in the Context of Quantum Probability
(2020) Lee, Yat-Ning Paul; Baecher, Gregory B; Civil Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
Conventional risk analysis and assessment tools rely on the use of probability to represent and quantify uncertainties. Modeling complex engineering problems with pure probabilistic approach can encounter challenges, particularly in cases where contextual knowledge and information are needed to define probability distributions or models. For the study and assessment of risks associated with complex engineering systems, researchers have been exploring augmentation of pure probabilistic techniques with alternative, non-fully, or imprecise probabilistic techniques to represent uncertainties. This exploratory research applies an alternative probability theory, quantum probability and the associated tools of quantum mechanics, to investigate their usefulness as a risk analysis and assessment tool for engineering problems. In particular, we investigate the application of the quantum framework to study complex engineering systems where the tracking of states and contextual knowledge can be a challenge. This study attempts to gain insights into the treatment of uncertainty, to explore the theoretical implication of an integrated framework for the treatment of aleatory and epistemic uncertainties, and to evaluate the use of quantum probability to improve the fidelity and robustness of risk system models and risk analysis techniques.
COMBINED ROBUST OPTIMAL DESIGN, PATH AND MOTION PLANNING FOR UNMANNED AERIAL VEHICLE SYSTEMS SUBJECT TO UNCERTAINTY
(2019) Rudnick-Cohen, Eliot; Azarm, Shapour; Herrmann, Jeffrey W; Mechanical Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
Unmanned system performance depends heavily on both how the system is planned to be operated and the design of the unmanned system, both of which can be heavily impacted by uncertainty. This dissertation presents methods for simultaneously optimizing both of these aspects of an unmanned system when subject to uncertainty. This simultaneous optimization under uncertainty of unmanned system design and planning is demonstrated in the context of optimizing the design and flight path of an unmanned aerial vehicle (UAV) subject to an unknown set of wind conditions. This dissertation explores optimizing the path of the UAV down to the level of determining flight trajectories accounting for the UAVs dynamics (motion planning) while simultaneously optimizing design. Uncertainty is considered from the robust (no probability distribution known) standpoint, with the capability to account for a general set of uncertain parameters that affects the UAVs performance. New methods are investigated for solving motion planning problems for UAVs, which are applied to the problem of mitigating the risk posed by UAVs flying over inhabited areas. A new approach to solving robust optimization problems is developed, which uses a combination of random sampling and worst case analysis. The new robust optimization approach is shown to efficiently solve robust optimization problems, even when existing robust optimization methods would fail. A new approach for robust optimal motion planning that considers a “black-box” uncertainty model is developed based off the new robust optimization approach. The new robust motion planning approach is shown to perform better under uncertainty than methods which do not use a “black-box” uncertainty model. A new method is developed for solving design and path planning optimization problems for unmanned systems with discrete (graph-based) path representations, which is then extended to work on motion planning problems. This design and motion planning approach is used within the new robust optimization approach to solve a robust design and motion planning optimization problem for a UAV. Results are presented comparing these methods against a design study using a DOE, which show that the proposed methods can be less computationally expensive than existing methods for design and motion planning problems.
Development of Risk-Based Measurements and Metrics for Sustainability Quantification of Manufactured and Constructed Systems
(2018) Webb, David Harry; Ayyub, Bilal M; Reliability Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
Sustainability has been an important topic of study for several decades; however, its importance has escalated with the signing of the Paris Agreement. One issue that has always hindered implementing sustainability research in practice has been the difficulty in measuring performance. While methods such as life-cycle assessment are available to enable a comparison with alternatives, sustainable performance cannot be related to larger environmental goals. Additionally, such methods often omit uncertainty considerations. The proposed research herein provides foundational measurement science and metrics to bridge the gap between the theories of sustainability and the application. The metrics enable tracking of measurable progress in all aspects of sustainability within a risk-based framework. This dissertation opens by reviewing and analyzing the literature on sustainability definitions and existing metrics in order to determine the current state of the practice, and to inform the development of the proposed metrics. Next, in order to demonstrate the capacity of risk-based approaches in measuring sustainability performance, a methodology is proposed to calculate the probability of a structure or product meeting sustainability requirements. Last, the methodology is validated using the National Institute of Standards and Technology’s Building Industry Reporting and Design for Sustainability. The validation procedure demonstrated that the methodology was capable of reproducing results from a well-vetted database. The proposed methodology serves as the first step in a “sustainability reliability” metric that is practical, accurate and comprehensive in its coverage.
Essays on Organizational Choices under Uncertainty
(2017) Sharma, Siddharth; Beckman, Christine; Chung, Wilbur; Business and Management: Management & Organization; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
Environmental uncertainty has been widely studied by organizational theorists and strategy scholars. In this dissertation, I aim to contribute towards a further understanding of the implications of environmental uncertainty on organizational choices. I develop a general framework, across the two chapters, which links the effect of uncertainty on organizational choices, mediated by changes in the competitive landscape. In my two chapters, I look at different types of uncertainty namely, state uncertainty and effect uncertainty. I explore how these types of uncertainties impact the competitive landscape either by compressing performance difference between organizations and changing the viability of positions on the landscape respectively. As a consequence of the changing landscape, I study the strategic behavior response of organizations as they engage in risk-taking or repositioning. I test my theoretical predictions across two interesting empirical constructs of Formula 1 car racing and the Consumer Electronics Show. Also, I also employ the use of a simulation model in my second chapter to supplement my empirical context.
Essays on Uncertainty, Imperfect Information, and Investment Dynamics
(2016) Jia, Dun; Aruoba, Boragan; Stevens, Luminita; Economics; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)
Understanding how imperfect information affects firms' investment decision helps answer important questions in economics, such as how we may better measure economic uncertainty; how firms' forecasts would affect their decision-making when their beliefs are not backed by economic fundamentals; and how important are the business cycle impacts of changes in firms' productivity uncertainty in an environment of incomplete information. This dissertation provides a synthetic answer to all these questions, both empirically and theoretically. The first chapter, provides empirical evidence to demonstrate that survey-based forecast dispersion identifies a distinctive type of second moment shocks different from the canonical volatility shocks to productivity, i.e. uncertainty shocks. Such forecast disagreement disturbances can affect the distribution of firm-level beliefs regardless of whether or not belief changes are backed by changes in economic fundamentals. At the aggregate level, innovations that increase the dispersion of firms' forecasts lead to persistent declines in aggregate investment and output, which are followed by a slow recovery. On the contrary, the larger dispersion of future firm-specific productivity innovations, the standard way to measure economic uncertainty, delivers the ``wait and see" effect, such that aggregate investment experiences a sharp decline, followed by a quick rebound, and then overshoots. At the firm level, data uncovers that more productive firms increase investments given rises in productivity dispersion for the future, whereas investments drop when firms disagree more about the well-being of their future business conditions. These findings challenge the view that the dispersion of the firms' heterogeneous beliefs captures the concept of economic uncertainty, defined by a model of uncertainty shocks. The second chapter presents a general equilibrium model of heterogeneous firms subject to the real productivity uncertainty shocks and informational disagreement shocks. As firms cannot perfectly disentangle aggregate from idiosyncratic productivity because of imperfect information, information quality thus drives the wedge of difference between the unobserved productivity fundamentals, and the firms' beliefs about how productive they are. Distribution of the firms' beliefs is no longer perfectly aligned with the distribution of firm-level productivity across firms. This model not only explains why, at the macro and micro level, disagreement shocks are different from uncertainty shocks, as documented in Chapter 1, but helps reconcile a key challenge faced by the standard framework to study economic uncertainty: a trade-off between sizable business cycle effects due to changes in uncertainty, and the right amount of pro-cyclicality of firm-level investment rate dispersion, as measured by its correlation with the output cycles.

Theses and Dissertations from UMD

Browse

Filters

Settings

Sort By

Results per page

Search Results