Systematic Design of Control Barrier Functions for Safety-Critical Systems

University Of Colorado At Boulder, Boulder CO

Investigators

Abstract

This grant will support research that will contribute new knowledge related to the design of reliable constrained control laws for autonomous systems. Many engineered physical systems are subject to restrictions and constraints that must be satisfied to ensure desirable behavior. When maximizing performance, it is often necessary to push such systems to their limits while being certain that they will not exit a safe and prescribed operating regime. This can be achieved using constrained control, which is a key enabling technology that allows systems to operate at their full potential, while guaranteeing safety. This award supports fundamental research to provide needed knowledge to systematically generate constrained control laws that are guaranteed to enforce safety requirements while also trying to boost performance. Results from this research will therefore benefit the U.S. economy and society. This research project involves expertise in several branches of applied mathematics, including control theory, set theory, and optimization. Control barrier functions have shown great promise for the deployment of computationally efficient constrained controllers. Unfortunately, existing literature does not provide systematic methods for generating control barrier functions, which has led to the deployment of constrained controllers that are unable to formally guarantee constraint satisfaction. To address this gap, this research effort aims to fill the knowledge gap by adapting invariance-based methods found in the reference governor literature to systematically generate control barrier functions. The team of researchers will investigate both continuous and discrete-time systems, address common robustness concerns, and validate the results on robotic applications. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.