Longitudinal velocity car-following control and optimization based on distributed model predictive control: Modeling, stability analysis and joint simulation

To better depict the car-following scenario of automated vehicles platoon and optimize the car-following performance, this paper designs a distributed model predictive control strategy based on the optimal velocity model (OVM) and longitudinal control (namely, OVLC-DMPC). Firstly, considering the coordinated control of velocity and acceleration, the optimal velocity longitudinal control model for platoons of connected and automated vehicles (CAVs) is designed, and a discrete-time system model is obtained through linearization and discretization. Secondly, the distributed model predictive control (DMPC) method is introduced, the car-following platoon is regarded as multiple independent systems for model predictive control (MPC), an upper-layer model predictive controller is established, the cost function and system control input constraints are designed, and the car-following velocity and headway of the vehicle are optimized to obtain the optimal model input. Thirdly, the eigenvalue determination of the discrete system and the Lyapunov method are used to conduct a theoretical analysis of the system’s stability. Finally, under the PLF communication topology, a joint simulation experiment of the system model is carried out based on CarSim and MATLAB/Simulink. The lower-layer controller is used to convert the output of the model predictive controller into throttle opening/braking pressure, thereby realizing the velocity control of the CarSim vehicle model. The system’s velocity tracking performance and anti-disturbance performance are evaluated through comparative experiments. The results show that under the OVLC-DMPC control strategy, fast-tracking of the vehicle platoon can be achieved, thereby enhancing the vehicle platoon's stability and the road's traffic capacity.