Longitudinal control of connected autonomous vehicles platoons with neuro sliding mode controller

Connected autonomous vehicles (CAV) have recently become an important research topic due to their potential to significantly affect road transport such as traffic accidents, traffic congestion and ensuring traffic safety. Therefore in this study, a neuro-sliding mode controller (NSMC) is designed to provide longitudinal control of connected autonomous vehicles platoons. First, the formulation of the longitudinal dynamics of the CAV platoon is obtained. Afterwards, the neuro-sliding mode controller design was carried out under uncertain dynamics that change over time by formulating platoon control targets in order to ensure the string stability of the vehicle platoon and to prevent rear end collisions. The accuracy of the designed controller is demonstrated by simulation studies after theoretical verification with Lyapunov stability criteria. As a result, the nonlinear mathematical model of CAV was created by taking into account external disturbances and parameter uncertainties, and the NSMC controller design was implemented under uncertain dynamics that change over time. With the help of neural networks in the designed controller structure, unknown system dynamics were learned, and the desired inter-vehicle distances were achieved using a sliding mode controller, and longitudinal string stability in the vehicle platoon was successfully achieved.