Show simple item record

dc.contributor.advisorMavris, Dimitri N.
dc.contributor.authorChoi, Youngjun
dc.date.accessioned2016-08-22T12:23:47Z
dc.date.available2016-08-22T12:23:47Z
dc.date.created2016-08
dc.date.issued2016-07-18
dc.date.submittedAugust 2016
dc.identifier.urihttp://hdl.handle.net/1853/55637
dc.description.abstractThe integration of Unmanned Aircraft Systems in the National Airspace System (UASNAS) problem has received much attention because of the growing number of variety of mission types and the rapid growth of UAS market. Among the many challenging UASNAS problems, separation assurance is considered to be particularly complex, having many interactions among the elements in different levels of abstraction and coupling effects between the different disciplinary domains. In order to explore the separation assurance problem, an analytic model should capture diverse operational scenarios, vehicle dynamics, and subsystem functions such as sensor/surveillance, control, navigation and communications. This has major implications on the analytic model requirements, especially in regard to modeling scope, resolution (or fidelity), and computational expense. The objective of this thesis is to formulate and demonstrate improvements in modeling and simulation of fully integrated UAS to enable systems analysis across the levels of abstraction and multiple disciplines. This work also quantitatively characterizes collision avoidance as a critical element of separation assurance in terms of system behaviors across the levels of abstraction and multiple disciplines. To address these objectives, this thesis contributes to four areas: (1) a statistical gain-scheduling method to improve computational efficiency without a loss of accuracy or fidelity, (2) a hybrid collision avoidance algorithm using a machine learning technique that improves computational runtime as well as optimal trajectory cost, (3) a two-layer obstacle avoidance algorithm for a multi-obstacle environment, (4) a rapid, data-driven and grid-based urban modeling methodology using airborne LiDAR sources. The proposed modeling and simulation capability provides insights into the interaction between system of systems, systems, and subsystems that cannot be characterized by a conventional modeling and simulation environment. To illustrate the collision avoidance problem, this thesis examines the navigation of a fixed wing UAV in a dense urban environment.
dc.format.mimetypeapplication/pdf
dc.language.isoen_US
dc.publisherGeorgia Institute of Technology
dc.subjectGain-scheduling method
dc.subjectCollision avoidance algorithm
dc.subjectMachine learning algorithm
dc.subjectUrban modeling methodology
dc.titleA framework for modeling and simulation of control, navigation, and surveillance for unmanned aircraft separation assurance
dc.typeDissertation
dc.description.degreePh.D.
dc.contributor.departmentAerospace Engineering
thesis.degree.levelDoctoral
dc.contributor.committeeMemberSchrage, Daniel P.
dc.contributor.committeeMemberFeron, Eric
dc.contributor.committeeMemberJimenez, Hernando
dc.contributor.committeeMemberValasek, John
dc.date.updated2016-08-22T12:23:47Z


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record