Optimal Control of Hybrid Systems
Hybrid systems can accurately model a large number of systems in many application domains where traditional non-linear systems can not. Since they are generally not as well understood, the options to control hybrid systems are usually more limited. Recently, there has been an enormous interest to use optimal control for these systems, and many results have shown that that optimal control is indeed a good method to control these systems. However, since optimal control is generally open-loop, the control is prone to measurement noise or disturbances. The goal of this research project is to close this loop, and provides methods to introduce feedback into the control loop and increase robustness of the control technique.
Real-Time Optimal Control of UAVs
Current state-of-the-art when controlling and coordinating multiple unmanned aerial vehicles (UAVs) requires multiple operators for controlling a single UAV. Through this program, this relationship will be inverted in that a single pilot will be enabled to effectively control and coordinate up to three UAVs simultaneously. This will allow the pilots to operate much more effectively, and will moreover enable the pilots to coordinate and exploit capability synergies between different UAVs to accomplish the mission objectives more effectively.
The key technology that will be developed for this is a pilot decision aids system in which the pilot (the operator) can choose to control one UAV directly and let the remaining UAVs organize themselves autonomously. If no such UAV is selected, the vehicles will operate in a fully autonomous mode, based on the predefined mission specifications. Moreover, the system will be able to predict what effect the selection of a particular vehicle as leader UAV will have on the overall mission, and the performance of the system will be demonstrated against an operational surveillance scenario in a 3D simulation environment.
- X.C. Ding, A. Schild, M. Egerstedt, and J. Lunze, Real-Time Optimal feedback Control of Switched Autonomous Systems. Submitted to the IFAC Conference on Analysis and Design of Hybrid Systems, Zaragoza, Spain, Sept. 2009.
- A. Schild, X.C. Ding and M. Egerstedt, Design of optimal switching surfaces for switched autonomous systems. IEEE Conference on Decision and Control, Shanghai, China, Dec. 2009. In submission.
- D. Ding, Y. Wardi, and M. Egerstedt. Adaptive Optimal Timing Control of Hybrid Systems. Mathematical Theory of Networks and Systems, Blacksburg, VA, July 2008. To appear.
- D. Ding, Y. Wardi, D. Taylor, and M. Egerstedt. Optimization of Switched-Mode Systems with Switching Costs. American Control Conference, Seattle, WA, June 2008. To appear.
- Y. Wardi, X.C. Ding, M. Egerstedt, and S. Azuma. On-Line Optimization of Switched-Mode Systems: Algorithms and Convergence Properties. IEEE Conference on Decision and Control, New Orleans, LA, Dec. 2007.
- Y. Wardi, X. Ding, and S. Azuma. On-Line Optimization of
Switched-Mode Hybrid Dynamical Systems. Intl. Conf. on Hybrid Systems: Computation
and Control, Pisa, Italy, April 3-5, 2007.
National Science Foundation