will propose a doctoral thesis entitled,
Comparing and Improving VTOL Aircraft Designs with Model Predictive Control
On
Friday, December 13 at 9:00 a.m.
Weber 200
Teams Link
Abstract
VTOL aircraft are complex, difficult to control, and push the limits of energy and propulsion technology. Their CONcepts of OperationS (CONOPS) are also among the most demanding in terms of performance, control, and safety. A variety of vehicle concept designs and control topologies offer diverse advantages to cope with these challenges. It is important to choose the best concept design and control topology for any particular CONOPS, but it is difficult to establish a fair basis of comparison. For example, is a vehicle with more rotors safer? The answer depends in part on whether the vehicle is still sufficiently controllable after losing one rotor. If so, then it is arguably safer in terms of controllability. If not, then it may be less safe due to the additional single points of failure. What if an aircraft could be made lighter by changing certain actuators’ type, size, or location? Any such design change would be valid only if control performance is still sufficient after the change. To answer such questions, a VTOL design process demands a means for evaluating control performance. The analysis must optimize the control for each case to enable different vehicle concept designs and control topologies to compete fairly. Furthermore, this control optimization must consider extreme operating points, since these conditions are what typically size actuators. Classical control sizing strategies do not include control optimization and do not consider certain important control performance metrics. Newer methodologies include control optimization but do not incorporate constraints to account for extreme operating points. This thesis proposes a framework leveraging Model Predictive Control (MPC) as an ideal control architecture to enable control performance analysis within the VTOL aircraft design process. The research will also explore actuator prioritization and mixing within the MPC framework to answer certain practical concerns such as upset recovery and computational resource consumption. The thesis will also explore the sensitivity of control performance to control optimization objective functions such as energy usage, ride quality, and community noise. The framework will be applied to RAVEN, an innovative multi-tiltrotor electric VTOL (eVTOL) aircraft, to demonstrate feasibility of the framework for systems of such complexity. Different control architectures for RAVEN will be implemented and compared in the hover condition to demonstrate control analysis via MPC in the VTOL design process.
Committee