Guaranteed Vehicle Safety Control Using Control-Dependent Barrier Functions

Abstract

How to guarantee the safety of autonomous ground vehicles is still a significant challenge, although different active safety control systems have been developed, including techniques based on controlled invariant sets and control barrier functions (CBF) that have been proven to guarantee the safety of dynamic systems. To apply these techniques to vehicle safety control, we regarded an estimated lateral stability region of a vehicle as a controlled invariant set. However, the stability region was found to vary with respect to the system control input, which is normally the steering angle. Therefore, the existing definition of the controlled invariant set, which is independent of control inputs, may not be applicable. In this paper, the definition of a controlled invariant set is extended to a control-dependent invariant set, so that the controlled invariant set can vary with control inputs. Based on the extended definition, a new invariance condition using control-dependent barrier function (CDBF) is proposed and applied to the guaranteed vehicle safety control problem. Finally, the guaranteed vehicle safety control is demonstrated by simulation results.