Research On Lateral And Longitudinal Coordinated Control Method For Autonomous Vehicles

With the development of network communication,artificial intelligence and control technology,autonomous driving technology has attracted much attention in the field of vehicle.As the final product of automobile intelligence,safety is the eternal theme in the development process of autonomous vehicle.Efficient and accurate lateral and longitudinal motion control is one of the necessary conditions for autonomous vehicle to improve safety.In the context of vehicle intellectualization and frequent traffic accidents,improving the motion control performance of autonomous vehicles is helpful to ensure the safety,convenience and comfort of travel.In order to further improve the control accuracy of lateral and longitudinal motion of autonomous vehicles at high speed,this paper designs a lateral and longitudinal coordinated control method for autonomous vehicles.The main work and innovation are as follows:(1)Design of path planning algorithm based on driving hazard field.On the basis of the field theory,combined with the relevant factors affecting the driving safety during the driving process,the potential energy field model of static obstacle,kinetic energy field model of dynamic obstacle and potential energy model of lane line considering the speed were established respectively to describe the influence of the external environment on the controlled vehicle during the driving process.On this basis,the driving hazard field model was established to finally realize the speed and path planning of vehicles in three scenarios:lane maintenance,stable following and lane change overtaking.(2)Design of lateral controller based on variable time domain model predictive control.Firstly,the 3-DOF vehicle dynamics model was established,and based on this model predictive control algorithm model was established.As the parameters of the model predictive control algorithm are derived from engineering experience,there are some problems such as poor adaptability to changes in working conditions.To solve this problem,a lateral motion controller based on variable time domain model predictive control was designed to improve the control stability and accuracy of the controlled vehicle under variable speed conditions.Finally,the co-simulation platform of Car Sim and Matlab/Simulink was built,and the simulation was carried out under the condition of double line change.By comparison,the proposed lateral controller based on variable time domain model predictive control can realize lateral motion control with better control stability and accuracy than the fixed time domain controller and variable predictive time domain controller.(3)Design of a layered longitudinal controller considering the motion state of other vehicles.Aiming at the longitudinal motion control of the vehicle,the layered structure was adopted to design the longitudinal motion controller.The upper controller uses the model predictive control algorithm to calculate the expected acceleration according to the motion state of the vehicle in front.The lower controller uses the vehicle inverse longitudinal dynamics model to convert the expected acceleration into throttle opening and braking pressure.In addition,drive/brake switching logic was added to avoid damage to the vehicle system.Through Car Sim and Matlab/Simulink co-simulation platform verification,The longitudinal controller can realize the real-time planning and accurate tracking of the speed according to the motion state of other vehicles.It can ensure the driving safety and ride comfort.(4)Design of lateral and longitudinal coordination controller with replanning function.In order to further improve the precision of the motion control of the autonomous vehicle,the lateral and longitudinal controllers were coupled with the speed as the coupling point,and on this basis,a replanning system was designed according to the driving danger field,and finally a lateral and longitudinal coordination controller with replanning function was formed.In the high-speed curve overtaking condition,through the Car Sim and Matlab/Simulink co-simulation platform simulation verification,the controller can complete the real-time local path planning according to the motion state and space position of other vehicles,and complete the motion control of controlled vehicles according to the replanning path.Moreover,it has better control stability,track tracking accuracy and speed tracking accuracy in vehicle motion control.