On the robustness of hybrid control systems to measurement noise and actuator disturbances
For control engineers designing hybrid systems, the paper identifies a critical gap in robustness theory and offers a stronger definition to ensure practical implementations are robust.
The paper analyzes robustness of hybrid control systems to measurement noise and actuator disturbances, showing that existing robustness definitions fail for practical implementations. It proposes a new concept of 'strong robustness' that guarantees robustness of jumping-first and flowing-first implementations, and provides a sufficient condition based on hybrid relaxation.
Robustness of hybrid control systems to measurement noise, actuator disturbances, and more generally perturbations, is analyzed. The relationship between the robustness of a hybrid control system and of its implementations is emphasized. Firstly, a formal definition of implementation of a hybrid control system is provided, based on the uniqueness of the solutions. Then, two examples are analyzed in detail, showing how the previously developed robustness property fails to guarantee that the implementations, necessarily used in control practice, are also robust. A new concept of strong robustness is proposed, which guarantees that at least jumping-first and flowing-first implementations are robust when the hybrid control system is strongly robust. In addition, we provide a sufficient condition for strong robustness based on the previously developed hybrid relaxation results.