Research And Implementation Of Fast Tracking Algorithm For Autonomous Vehicle Trajectory Based On Improved MPC

In recent years,Autonomous Vehicle,as a leading field of artificial intelligence to infiltrate into traditional skilled industries,is showing explosive growth in the world,especially in China and the United States.Trajectory tracking control technology is at the end of the driverless technology chain,carrying the intelligent behavior of the main system.It is a bridge between vehicle and other modules,supporting and realizing the whole algorithm architecture.The purpose of this paper is to study the application of model predictive control(Model Predict Control,MPC)technology in autonomous vehicle motion control(Motion Control).The main contents of this paper are as follows:(1)Passenger vehicle has the characteristics of strong nonlinearity and high complexity.How to build a low dimensional and accurate vehicle mathematical model is a difficult problem that has been puzzled by vehicle motion control and stability control scholars.First of all,this paper comprehensively evaluates the common vehicle mathematical models.Combined with the specific problems of vehicle lateral stability,a two wheel vehicle dynamics model with low complexity and lateral stability control and trajectory tracking control is improved.(2)The highly nonlinear vehicle model will also bring difficulties to the stability of the model analysis,this paper ignores the suspension system of the vehicle lateral stability problems in the horizontal plane,using the vehicle dynamics model skillfully uses the yaw rate(Yaw Rate)and lateral angle(Side Slip Angle)the stability of phase plane analysis the vehicle is composed of,and summarizes the classification of stable region and unstable region of the law,put forward a method of packet line constraint to adjust the vehicle lateral stability.(3)Studied to solve the vehicle kinematics model of the design of MPC controller based on the tracking problem of autonomous vehicle trajectory,designed a vehicle dynamics model of the improved MPC control system and the integration of solving the stability problem of packet line constraint based on the control system has the trajectory tracking and stability control of two parts.In order to meet the real time of the engineeringapplication,the third party solver is introduced to realize the high speed solution of the MPC problem.(4)Build a Carsim/Simulink/Matlab simulation platform,design a variety of scene simulation test design of control accuracy and reliability of the system,the simulation results verify the high real-time control system designed in the engineering application,the trajectory tracking performance and robustness in the lateral stability of the effectiveness.