Simulation Development, Path Planning, and Linear Quadratic Gaussian Precision Landing for Unmanned Traffic Management Operations

Abstract

Unmanned Air Systems (UAS) are heavily utilized in various missions for both military and consumer applications. Because of their immense popularity, the Federal Aviation Administration (FAA) has derived a notional architecture for an Unmanned Traf- fic Management (UTM) system to regulate the airspace for UAS operations. In addition, NASA has conducted research on the Extensible Traffic Management, where the airspace will be shared by both manned and unmanned systems [16]. This thesis addresses the issue of the need for a High Fidelity Simulation Environment to test the interaction between simulated UAS with UTM and its Universal Service Suppliers (USS). The framework proposed provides extensibility for additions of different USS for testing protocols. This thesis also contributes a hierarchical multi-UAS path-finding algorithm that can be adjusted for different topologies as well as mimics standard route planning methods conducted in manned aircraft operations. To test this simulation 30,000 total Monte Carlo simulations were conducted for 10 to 100 UAS operators, with different prioritization techniques. In addition, the method was tested in the High Fidelity simulation. Last this thesis provides a control and guidance law for tracking and precision landing of quadcopters for future applications of USS and industry applications, utilizing AprilTags with a Linear Quadratic Gaussian (LQG) law. Tests were conducted with the High Fidelity framework developed. Results indicate that the guidance and control law is robust from gust disturbances injected in the simulation realm.