Uncertainty Quantification for Robust Control Design of Smart Material Systems

The objective in robust control design is to provide mechanisms to achieve tracking or stabilization objectives in the presence of unmodeled dynamics. This is usually achieved by assuming worst case model discrepancies which can significantly degrade control authority if the uncertainty bounds are overly conservative. In this paper, we use uncertainty quantification techniques to construct densities for control outputs that can be used to derive optimal robust control designs. We illustrate the performance of these techniques in the context of systems with smart material actuators and sensors.