CAREER: Breakthroughs in Dynamical Modeling and Control for Reduction of Catastrophic Aviation Accidents

University Of Washington, Seattle WA

Investigators

Abstract

This Faculty Early Career Development (CAREER) project will support fundamental research in control technology to safely land aircraft in certain adverse situations. Specifically, a significant number of aircraft accidents occur when the primary control system is partly or completely lost due to structural or hydraulic failures. Due to the critical role played by the aviation industry in transportation, tourism, and economic growth, it is imperative to reduce occurrences that lead to loss of life or property.The new mathematical foundations laid down in this research will enable control and recovery in highly adverse failure conditions using inherent controllability properties of multiple jet engine configurations. The high pilot workload in emergency situations coupled with slow engine response makes precise control in general very difficult to accomplish, therefore this research will be used as the basis for systems to assist flight crew of all skill levels to successfully carry out emergency landings. Once the new technology is tested in simulation under various scenarios, it will be integrated in pilot training modules that will further help mitigate accidents due to inappropriate crew action. Additionally, the findings of this research will expand the operational envelope of unmanned aerial vehicles that are becoming an increasingly important part of the national airspace. Multiple engines on a commercial jet can be used for directional control under emergency situations. However, the engine response (time constant of 4 seconds or more) is typically very slow compared to the response of primary controls (with time constant of 0.35 seconds). This significant lag often leads to a high gain control signal that exceeds physical limits of the controller. This research addresses this scientific barrier by deriving a new dynamical modeling procedure that captures the effects of significant lag between the commanded control signal and the actual applied control signal on nonlinear systems using concepts from singular perturbation theory. Using these new mathematical representations the research team will create an improved performance, stable and robust control architecture. One major benefit of the derived controller is that the feedback solution is independent of the large time constant. The research team will also integrate the controllers with pilot feedback to strengthen pilot/operator training for automated aircraft and to create enhanced training procedures and pilot interfaces.