(Georgia Institute of Technology, 2018-04-30)
Degraw, Christopher F.
Given the plans for satellite mega-constellations, there is a lack of rigorously tested operations and control methods for constellations larger than 30 to 50 spacecraft. The purpose of this thesis is to propose the principles behind a robust, modular, and scalable system able to provide software-in-the-loop (SWIL) and hardware-in-the-loop (HWIL) simulation capabilities for the advancement of formation and constellation system Technology Readiness Levels (TRL). Additionally, this thesis will develop a first generation system demonstrating these principles called Constellation Simulation on a Massive Scale, or COSMoS. The preliminary goals of COSMoS are to 1) simulate multiple or more satellites in a constellation to demonstrate scalable capability; and 2) connect to external hardware devices in real-time to demonstrate HWIL capability. The simulation framework behind COSMoS is the Multi-Agent Distributed Network Simulator, or MADNS. MADNS is a real-time hardware-in-the-loop (RT-HWIL) simulator capable of communicating with independent agents and external hardware and software elements. This framework will encapsulate the COSMoS simulation but will be designed to work with any multi-agent network simulation designed within the constraints of the MADNS API. This thesis will show the results of the preliminary development of both MADNS and COSMoS and will present a direction for the further development of both a satellite constellation simulator and general real-time hardware-in-the-loop simulators.
(Georgia Institute of Technology, 2018-04-26)
Greenhill, Andrew
This thesis investigates the effects of two specific sensor limitations in enhanced flight vision systems (EFVS) on general aviation pilot performance during approach and landing: sensor range and EFVS portrayal of runway markings. The background section of this thesis describes current sensor technologies with EFVS: millimeter wave radar, forward-looking infrared, and light detection and ranging (LiDAR). In addition, the connections between pilot tasks, information requirements, visual cues and information processing level are identified. These connections show how limitations of sensor technologies could affect pilot performance. These effects were then assessed in a fixed base flight simulator of a general aviation aircraft with an EFVS system. The sensor range and portrayal of runway markings was varied while measuring pilot performance. Pilot performance during approach was measured according to FAA instrument certification standards. Landing performance was measured using standards taught during private pilot training. The results show that pilot performance in tracking an instrument approach is negatively affected by reductions in EFVS sensor range, while the vertical speed and distance from centerline had exceedances beyond acceptable standards when the EFVS did not portray runway markings. These results identify the key minimum specifications of EFVS sensor range and ability to portray runway markings for their implementation in general aviation.