Resilient Neural-Sliding-Mode Distance Regulation in Merging Control of Heterogeneous Connected Automated Vehicles Subject to Deception Attacks

Resilient distance regulation process is significantly essential in intelligent transportation systems with connected automated vehicles (CAVs), due to vulnerability of communication links within the vehicles of a platoon. On the other hand, time varying distance regulation process is usually ignored in adaptive cruise control procedures, which yields to be impossible for other vehicles on the right lane or the ramp to merge to the platoon. So, the main contribution of this paper is to design an adaptive resilient neural-sliding-mode cruise control with distance regulation procedure to make the following vehicles to track the leader's velocity and acceleration profile, while keeping a time-varying safe distance between each two consecutive vehicles, even when another vehicle seeks to merge into the platoon. To this end, at first an adaptive neural sliding mode control procedure along with a variable structure virtual disturbance observer is designed to provide a safe and smooth time-varying distance regulation process for the vehicle receiving the merging signal as well as maintaining a predefined distance between other two vehicles, while there exists cyber-attacks, external disturbances, and unknown nonlinearity in the system. Furthermore, it is assured that the followers track the leader's velocity and acceleration profile in a leader-to-all topology. To the authors’ knowledge, this is the first time that a resilient nonlinear cruise control procedure is proposed with distance regulation process for a platoon of nonlinear CAVs facing unknown nonlinearities and deception attacks. Finally, numerical results validate the effectiveness of the proposed procedure on achieving control objectives, resilience, and advantages compared to a recently proposed method.