Rule-Based Safety-Critical Control Design using Control Barrier Functions with Application to Autonomous Lane Change

Abstract

This paper develops a new control design for guaranteeing a vehicle's safety during lane change maneuvers in a complex traffic environment. The proposed method uses a finite state machine (FSM), where a quadratic program based optimization problem using control Lyapunov functions and control barrier functions (CLF-CBF-QP) is used to calculate the system's optimal inputs via rule-based control strategies. The FSM can make switches between different states automatically according to the command of driver and traffic environment, which makes the ego vehicle find a safe opportunity to do a collision-free lane change maneuver. By using a convex quadratic program, the controller can guarantee the system's safety at a high update frequency. A set of pre-designed typical lane change scenarios as well as randomly generated driving scenarios are simulated to show the performance of our controller.

Robust Safe Control Synthesis with Disturbance Observer-Based Control Barrier Functions

“…CBFs can be unified with stability and performance requirements, encoded by the time derivative of a control Lyapunov function (CLF), in an online quadratic program (CLF-CBF-QP) to ensure safety and control objectives simultaneously [2]. This optimization framework has been widely applied to multi-agent systems [3], autonomous driving [4], and wheeled robots [5] due to its computational efficiency. Although these applications guarantee optimization constraints for high-fidelity dynamical models, real-time control systems generally include unavoidable uncertainties and disturbances that might cause performance degradation, and in some cases, even lead to unsafe operations [6], [7].…”