A Real-Time Performance Recovery Framework for Vision-Based Control Systems

The demand for high performance of vision-based control systems is ever increasing. This article, thus, proposes, for the first time, a real-time performance recovery framework for the vision-based control of highly dynamic systems. The framework enables functionalized nominal controller and performance recovery, balanced image processing, and performance recovery, as well as enhanced real-time efficiency, which is based on the observer-based realization of the Youla parameterization of all stabilizing controllers. An online learning algorithm for real-time performance recovery against the unexpected performance degradation is developed as well. The verification of the proposed framework and algorithm promises their applications to classes of vision-based highly dynamic control systems.