An autonomous optimal safe prescribed performance tracking control strategy for homogeneous vehicular platoons encountering sudden obstacles

The optimal autonomous safe prescribed performance tracking control problem is studied for homogeneous vehicular platoon systems (VPSs) encountering sudden obstacles in this paper. It focuses on the situation where obstacles suddenly intrude into the driving lane of a vehicle platoon. A secure boundary protection method (SBPM) is used to design an autonomous safety tracking strategy for emergency obstacle avoidance, which includes the control design after obstacle removal. When an obstacle suddenly enters the lane and conflicts with the desired trajectory, the strategy can automatically construct two adjustable secure boundaries by using the SBPM, so that the VPS enters a safe region in advance to avoid collision. After the obstacle is removed, the platoon can autonomously track the original desired trajectory based on a new secure boundary self-adjustment law (SBSAL). In the controller design, a novel constraint control method, which is independent of the initial spacing errors between adjacent vehicles, is adopted based on a variable barrier function (VBF), and the string stability of the platoon is guaranteed. Specifically, only one collective performance constraint function is needed for all vehicles in this platoon. The reinforcement learning (RL) algorithm is utilised to optimise the control input of each vehicle in the VPS. Finally, the effectiveness of the proposed method is verified through simulations.