Aerial Refueling Simulator Validation Using Operational Experimentation and Response Surface Methods with Time Series Responses.

摘要

An important program in the Department of Defense is the KC-46 Supertanker. Dubbed the future of the Air Force's aerial refueling inventory, the KC-46 will replace dozens of ailing previous generation tanker aircraft. The Aerial Refueling Airplane Simulator Qualification document governs the methods by which Air Mobility Command validates its simulators, some of which will be KC-46 simulators in the near future. The methodology set forward in this thesis utilizes historical data of aircraft performance from similar air frames to gain statistical insight into the performance design space of the KC- 46. Leveraging this insight, the methodology provides through a framework for validation that uses classical experimental design principles as applied to time history responses such as found in aircraft performance measures. These principles guide the generation of response surfaces from real world flight test data that can then be used to validate flight training simulators using a point by point comparison or over an entire surface of points for a variety of different aerial refueling maneuvers. This work also supports the KC-46 Tanker program by proposing statistically efficient and cost conscious experimental designs for the KC-46 flight testing. This framework is demonstrated using flight testing data from the KC-135 Aerial Refueling Simulator Upgrade testing, and is part of an Office of the Secretary of Defense initiative to add increased statistical rigor to the Department of Defense test and evaluation enterprise and specifically the acquisition community.