Research On Collaborative Control Strategy Of Autonomous Vehicle Fleet

The cooperative driving of fleet can unify the management of vehicles with common driving purpose,which can alleviate traffic congestion and simplify traffic management,and become a hot spot in the research of intelligent vehicle-road cooperative system.However,due to the complexity and dynamic change of the driving traffic environment,it is difficult to realize the stable and safe driving of the fleet in the real traffic environment.This paper takes the fleet collaborative driving system as the research object,using the theory of composite system,establish a more collaborative driving car composite system architecture.Aiming at the coordination layer and the control layer,the cooperative driving state switching logic and the lateral and longitudinal controller of the fleet are designed.Meanwhile,the path and speed planner are designed to realize the stable and safe driving of the fleet in a more complicated traffic environment.The main research contents of this paper are as follows:Firstly,the research status of the architecture of the fleet cooperative driving system is analyzed,it is clear that the collaborative driving of the fleet meets the characteristics of the hybrid system.The advantages and disadvantages of the fleet longitudinal control strategy and the lateral and longitudinal control strategy are compared,and the current control strategy to achieve the fleet cooperative lane change is analyzed.Secondly,aiming at the control layer of the cooperative driving system of the fleet,the longitudinal structure model of the fleet was established.The model predictive control algorithm was used to establish the longitudinal upper controller of the fleet.Combined with the longitudinal dynamic characteristics of the vehicle,the engine torque characteristic curve and the relationship between the brake pressure and the brake torque were used to establish the longitudinal lower controller of the fleet.Prediction algorithm according to the characteristics of the vehicle acceleration,using the model,the design of vehicle longitudinal controller,according to the transverse vehicle longitudinal dynamics model,using the model prediction control algorithm,establish the vehicle lateral controller,the longitudinal controller and the lateral controller are coupled according to the longitudinal speed,exchange condition and speed and operating mode of obstacle avoidance path planning and design,The path of lane change adopts quintic polynomial function,the path of obstacle avoidance adopts double line shift function,and the speed planning adopts trapezoidal function.According to the characteristics of the motorcade,the lateral and longitudinal controller of the motorcade was designed.The leading vehicle carried out speed and path planning,followed the vehicle to realize the expected speed and spacing tracking,and independently carried out path planning and tracking.Thirdly,aiming at the coordination layer of the fleet cooperative driving system,the state switching logic is designed according to the characteristics of the fleet cooperative driving state,and the state switching logic is modeled in Stateflow by using the finite machine theory.The fleet cooperative driving state can be divided into leading-following,combination,splitting and lane changing.Finally,Truck Sim/Simulink co-simulation is used to verify that the designed coordination layer and control layer can be effectively controlled under typical operating conditions.