Safety Critical Control of Mixed-autonomy Traffic via a Single Autonomous Vehicle

Connected and automated vehicles (CAVs) have shown great potential in improving traffic throughput, stability, and safety in mixed-autonomy traffic. Many recent research studies focus on developing longitudinal control strategies for CAV s to achieve string stability of a heterogeneous vehicle platoon but the potential impact of these control strategies on safety has not been fully addressed. This paper proposed a safety-critical control strategy for CAV leading cruise control (LCC) framework in which there are both preceding and following human-driven vehicles (HDV s) for a single CA V. Considering the safety challenges brought by sudden acceleration and deceleration of the HDVs around the CAV, we introduced the notion of control barrier functions (CBF) to impart collision-free safety on a controlled LCC system. The safety of both HDVs and CAV are incorporated into the constraint-handling control design using a quadratic program, in which the actual CAV control input from a nominal input is penalized quadratically while the CBF safety constraints of control input guarantee the nonnegative spacing between vehicles. The simulation results were conducted to validate the proposed control strategy for two safety-critical scenarios.