Consolidated Control Barrier Functions: Synthesis and Online Verification via Adaptation under Input Constraints

In this paper, we develop a novel adaptation-based approach to constrained control design under multiple state and input constraints. Specifically, we introduce a method for synthesizing any number of time-varying candidate control barrier functions (CBF) into one consolidated CBF (C-CBF) candidate, and propose a predictor-corrector optimization-based adaptation law for the weights of the constituent constraint functions that certifies the C-CBF as valid for a class of nonlinear, control-affine systems. We prove this result by showing that the adapted weights are guaranteed to confer sufficient control authority to meet the new, adaptive C-CBF condition in perpetuity despite input constraints, which thereby permits its use in a quadratic program based control law. We then illustrate the performance of our controller on an academic example, and further highlight that it is successful even for constraint functions with higher or mixed relative-degree by simulating a reach-avoid problem for bicycle robots, which we use to demonstrate how our approach out-performs two baseline approaches.