UMD Theses and Dissertations
Permanent URI for this collectionhttp://hdl.handle.net/1903/3
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 given thesis/dissertation in DRUM.
More information is available at Theses and Dissertations at University of Maryland Libraries.
Browse
2 results
Search Results
Item ADAPTIVE SAMPLING METHODS FOR TESTING AUTONOMOUS SYSTEMS(2018) Mullins, Galen Edward; Gupta, Satyandra K; Mechanical Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)In this dissertation, I propose a software-in-the-loop testing architecture that uses adaptive sampling to generate test suites for intelligent systems based upon identifying transitions in high-level mission criteria. Simulation-based testing depends on the ability to intelligently create test-cases that reveal the greatest information about the performance of the system in the fewest number of runs. To this end, I focus on the discovery and analysis of performance boundaries. Locations in the testing space where a small change in the test configuration leads to large changes in the vehicle's behavior. These boundaries can be used to characterize the regions of stable performance and identify the critical factors that affect autonomous decision making software. By creating meta-models which predict the locations of these boundaries we can efficiently query the system and find informative test scenarios. These algorithms form the backbone of the Range Adversarial Planning Tool (RAPT): a software system used at naval testing facilities to identify the environmental triggers that will cause faults in the safety behavior of unmanned underwater vehicles (UUVs). This system was used to develop UUV field tests which were validated on a hardware platform at the Keyport Naval Testing Facility. The development of test cases from simulation to deployment in the field required new analytical tools. Tools that were capable of handling uncertainty in the vehicle's performance, and the ability to handle large datasets with high-dimensional outputs. This approach has also been applied to the generation of self-righting plans for unmanned ground vehicles (UGVs) using topological transition graphs. In order to create these graphs, I had to develop a set of manifold sampling and clustering algorithms which could identify paths through stable regions of the configuration space. Finally, I introduce an imitation learning approach for generating surrogate models of the target system's control policy. These surrogate agents can be used in place of the true autonomy to enable faster than real-time simulations. These novel tools for experimental design and behavioral modeling provide a new way of analyzing the performance of robotic and intelligent systems, and provide a designer with actionable feedback.Item OBSERVABILITY-BASED SAMPLING AND ESTIMATION OF FLOWFIELDS USING MULTI-SENSOR SYSTEMS(2014) DeVries, Levi David; Paley, Derek A; Aerospace Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)The long-term goal of this research is to optimize estimation of an unknown flowfield using an autonomous multi-vehicle or multi-sensor system. The specific research objective is to provide theoretically justified, nonlinear control, estimation, and optimization techniques enabling a group of sensors to coordinate their motion to target measurements that improve observability of the surrounding environment, even when the environment is unknown. Measures of observability provide an optimization metric for multi-agent control algorithms that avoid spatial regions of the domain prone to degraded or ill-conditioned estimation performance, thereby improving closed-loop control performance when estimated quantities are used in feedback control. The control, estimation, and optimization framework is applied to three applications of multi-agent flowfield sensing including (1) environmental sampling of strong flowfields using multiple autonomous unmanned vehicles, (2) wake sensing and observability-based optimal control for two-aircraft formation flight, and (3) bio-inspired flow sensing and control of an autonomous unmanned underwater vehicle. For environmental sampling, this dissertation presents an adaptive sampling algorithm steering a multi-vehicle system to sampling formations that improve flowfield observability while simultaneously estimating the flow for use in feedback control, even in strong flows where vehicle motion is hindered. The resulting closed-loop trajectories provide more informative measurements, improving estimation performance. For formation flight, this dissertation uses lifting-line theory to represent a two-aircraft formation and derives optimal control strategies steering the follower aircraft to a desired position relative to the leader while simultaneously optimizing the observability of the leader's relative position. The control algorithms guide the follower aircraft to a desired final position along trajectories that maintain adequate observability and avoid areas prone to estimator divergence. Toward bio-inspired flow sensing, this dissertation presents an observability-based sensor placement strategy optimizing measures of flowfield observability and derives dynamic output-feedback control algorithms autonomously steering an underwater vehicle to bio-inspired behavior using a multi-modal artificial lateral line. Beyond these applications, the broader impact of this research is a general framework for using observability to assess and optimize experimental design and nonlinear control and estimation performance.