Distributed safety-critical control for linear homogeneous vehicle platoons subject to actuator and communication delays

This paper introduces a novel two-layer safetycritical controller, designed for cooperative adaptive cruise control (CACC) of a vehicle platoon subject to both communication and input delays. The first layer of the controller, referred to as a nominal controller, aims to achieve state synchronization among vehicles in the platoon with respect to a virtual leader, employing a combination of a decentralized controller and a distributed observer. Clearly, applying the nominal controller to each vehicle would synchronize the states of all the vehicles in the platoon, causing the platoon to reduce to a single point. Therefore, a collision avoidance mechanism is implemented via control barrier functions (CBFs) and tunable input-to-state safe control barrier functions (TISSf-CBFs) through predictor feedback, to ensure safety, by keeping a particular spacing policy. The framework allows the inclusion of additional constraints such as velocity constraints to respect road regulations, acceleration constraints to prevent excessive accelerations or decelerations, and actuator constraints to ensure compliance with the physical limitations of the vehicles. In fact, using this two-layer framework, most unconstrained CACC designs could be modified, such that particular constraints are satisfied. This procedure allows the separation of internal stability analysis from the safety guarantees. In conclusion, the effectiveness of the proposed controller is validated through simulation, demonstrating that string stable behavior can be achieved with a relatively small time headway, despite the simultaneous presence of actuator and communication delays.