Distributed Model Predictive Control Of Longitudinal Driving Vehicle Platoon

Vehicle platoon technology is the key technology of intelligent networked vehicle cooperative control,which can effectively reduce the traffic accidents caused by human factors such as fatigue driving.In addition,the platoon technology can effectively reduce the distance between the vehicles and reduce the air resistance when the platoon drive at high speed,thus reducing the fuel consumption and exhaust emissions.Considering the safety,consistency and ride comfort of longitudinal vehicle platoon,traditional control algorithms are difficult to meet the requirements of vehicle platoon system.Therefore,this paper adopts the model predictive control method which can effectively deal with multi-objective and system constraints,and carries out simulation experiments on the longitudinal driving vehicle platoon system to verify the effectiveness of the algorithm.In this paper,the longitudinal dynamics of vehicles is firstly simplified,and then the state-space model of vehicle platoon is established according to the longitudinal kinematics model of vehicle,and the layered control scheme is adopted to realize platoon control.The upper layer uses the model predictive control to solve the expected acceleration,and the lower layer uses the feed-forward and feedback control method to track the expected acceleration given by the upper controller.Then,based on the principle of distributed model predictive control,the optimization problem of vehicle platoon is transformed into a quadratic programming problem with constraints,and the performance problems such as safety and ride comfort are transformed into constraints.A distributed model prediction controller based on Nash iteration is designed,which makes the whole vehicle platoon reach Nash equilibrium by multiple iterations and communication.Finally,the huge communication and calculation burden caused by multiple communication and iteration,as well as the string stability of vehicle platoon are solved.The "virtual" state trajectory is used to replace the multiple communication,and the constraint of ensuring the string stability is added into the optimization problem,so that when the vehicle platoon appears transient error,the workshop distance error will not increase with the propagation of the vehicle platoon,which is proved theoretically.The effectiveness of the proposed distributed model predictive control algorithm is verified by the Carsim/Matlab co-simulation.