Research On Intelligent Vehicle Trajectory Tracking Control Considering Model Uncertainties

Intelligent vehicles are the "present tense" of automotive development,which is of great significance in improving travel efficiency and traffic safety.Autonomous driving is the highest stage of the development of intelligent vehicle technology,and its trajectory tracking technology is the core component of autonomous driving technology.Therefore,this paper takes the intelligent vehicle as the research object,aiming at the problems of low tracking accuracy and poor handling stability of intelligent vehicles in complex driving environments caused by the uncertainties of the system model,the trajectory tracking control of intelligent vehicles is studied as follows:Firstly,the vehicle model is analyzed,the three-degree-of-freedom vehicle dynamics model and the Pacejka empirical tire model are established,and the mechanical properties of the tire are analyzed.In addition,the uncertainty of the system model is analyzed,and a vehicle uncertainty model considering external curvature disturbance,tire cornering stiffness uncertainty,and longitudinal vehicle speed time-varying is established.Secondly,ignoring the external disturbances and model parameter uncertainties,using the advantages of model predictive control to solve the multivariable constrained optimization problem,an intelligent vehicle lateral trajectory tracking controller considering multiple dynamic constraints is designed based on the nominal model to realize the active steering control of intelligent vehicle.The effectiveness of the designed controller is verified by different simulation conditions.What’s more,considering the problems of low tracking accuracy and poor stability of the designed active front wheel steering tracking controller under complex working conditions such as high-speed,ice and snow roads,considering the advantages that direct yaw moment control can improve the longitudinal dynamic performance of vehicles,an integrated controller of active front wheel steering and direct yaw moment considering model uncertainty is designed based on the theory of robust predictive control and hierarchical control architecture.In order to improve the real-time performance of the controller,the optimization solution of the upper controller is divided into two parts:on-line optimization and off-line calculation,which outputs the required front wheel angle and additional yaw moment;the bottom controller uses the distribution method based on the lowest tire utilization to solve the torque of four wheels.Finally,the two controllers are compared and analyzed through the Simulink/Carsim simulation platform.The simulation results show that the integrated controller based on robust model predictive control has better control performance,and the effectiveness of the integrated controller is verified on a Hardware-in-the-loop platform.