Computational Environment for the Development of an FAA-Compliant Level-6 Flight Training Device

Abstract
A flight simulator can successfully achieve its purpose only if equipped with adequate mathematical models of the aircraft, its sub-systems, and the environment. The US Federal Aviation Administration (FAA) has instituted stringent regulations to ensure that flight simulators used for pilot training reach desirable levels of accuracy and fidelity. The purpose of this thesis is to present the development and application of a design strategy and the computational environment associated with it for building an aircraft simulation model that meets the FAA regulations for flight simulator performance. The proposed methodology is based on using flight test data in combination with analytical modeling tools and heuristics.
The Simulink simulation environment within Matlab has been selected due to its recognized capabilities, flexibility, and portability. Several interactive computational tools have been developed to support the development. Flight test data of a business class jet is used for the purpose of this research effort. An important part of the proposed strategy consists of selecting the flight data and converting them into a usable format for Matlab/Simulink. Parameter identification techniques must then be applied at specific points in the flight envelope of the aircraft in order to create an accurate flight dynamics model. Two such modeling techniques, in time and frequency domains, are used within this project. Lookup tables for the stability and control derivatives are built based on the dynamic pressure. Tuning of the aerodynamic model is required to meet all FAA criteria. Once the FAA objective tests are completed, another more organic set of tests are conducted by pilots. The outcomes of these subjective tests are analyzed and additional tuning of the aerodynamic and dynamic model are performed accordingly. Eventually, compliance with both FAA objective and subjective tests is ensured and demonstrated through several tuning iterations.
Research Advisor: Dr. Mario Perhinschi
Committee Members: Dr. Marcello Napolitano and Dr. Larry Banta

Return to event listing.