Robust Control Barrier Functions with Set-Membership Estimation and Learning

Описание: Speaker:
Sze Zheng Yong (Department of Mechanical and Industrial Engineering, Northeastern University, Boston, USA)

Abstract:
Recent research in safety control has leveraged the availability of accurate models to detect impending safety violations and to intervene accordingly. However, there is often a mismatch between the models that are used for algorithm design and the real systems. Thus, this talk introduces robust control barrier functions for two classes of uncertain systems. For the first class of uncertain parametric control affine systems, methods based on mixed-monotone decomposition and robust optimization are presented, where the robust controlled invariance condition remains linear in the control inputs despite nonlinear and time-varying uncertainties. In addition, our design includes set-membership estimation based on interval observers to reduce the conservatism of the proposed robust control barrier functions. Next, a novel robust data-driven control barrier function is introduced for the class of uncertain continuous systems to guarantee robust safety despite worst-case realizations of generalization errors from prior data. In particular, under various continuity assumptions, we leverage set-membership learning approaches for learning guaranteed upper and lower bounds of an unknown function from the data set to obtain a safe input set for robust controlled invariance. Finally, the talk will conclude with a brief discussion of some future opportunities and challenges for guaranteeing safety of uncertain systems.