Research On Intelligent Vehicle Trajectory Tracking Control Strategy

Intelligent vehicle has the advantages of intelligent,networked,energy saving,environmental protection,safety and convenience,etc.It has a broad application prospect in intelligent transportation system and military field,and is the only way to realize vehicle intelligence.Intelligent vehicle is an intelligent motion platform which integrates computer technology,sensing technology,communication technology,artificial intelligence and decision control technology.It is a complex nonlinear system.As a key part of intelligent vehicle decision control,trajectory tracking has a complex coupling relationship with vehicle kinematics and dynamics.The precision and efficiency of vehicle tracking control algorithm are the key indexes to evaluate the performance of intelligent vehicle.This paper mainly studies the driving state parameter estimation and trajectory tracking control strategy of intelligent vehicles,builds the simulation environment of single shift,double shift and real road trajectory,builds the dynamics model based on the vehicle dynamics theory,improves the cubature Kalman filter algorithm based on adaptive noise estimation,and estimates the vehicle driving state parameters.Coupled MPC control strategy realizes precise control of intelligent vehicle tracking.The main contents are as follows:(1)Based on the vehicle dynamics theory,the intelligent vehicle dynamics model with three degrees of freedom including longitudinal,lateral and yaw was derived,and the model was built in Simulink.Coupled with the vehicle dynamics model based on Carsim,the cosimulation platform was constructed,which laid the foundation for the subsequent cosimulation analysis and control strategy verification.(2)The Cubature Kalman Filter based on adaptive Sage-Husa noise estimation is improved in order to improve the estimation accuracy of key parameters of intelligent vehicle running state and characterize the strong nonlinear characteristics of intelligent vehicle.And the vehicle’s transverse and longitudinal speed,yaw speed and centroid side Angle are estimated.Two driving conditions of single shift and double shift were constructed.The Extended Kalman Filter and the improved algorithm were verified and compared at different speeds.The results showed that the estimation accuracy of the two algorithms was similar at low speeds.Under high speed conditions,ACKF algorithm has better estimation accuracy than EKF algorithm.(3)Design the MPC control strategy for the vehicle horizontally and longitudinally.In lateral control,the lateral displacement error and yaw Angle error were taken as feedback signals,and the steering compensation Angle calculated by PID error compensator was superposed with the optimal steering Angle calculated by MPC to obtain the steering Angle of the control vehicle.In longitudinal control,the acceleration pedal and brake master cylinder pressure are controlled by MPC to control the speed of the vehicle.In the “Carsim+Simulink co-simulation” platform,the transverse and longitudinal control were simulated and verified respectively.The results show that the transverse error and yaw Angle error of the MPC controller are reduced by 51.4% and 17.7% respectively when tracking the single line moving trajectory.When tracking the double-shift track,the LQR controller fails at the speed of80km/h,and the lateral error and yaw Angle error of the MPC controller are reduced by 52.1%and 9.97% respectively.(4)Using Open Street Map website to export real map data and construct corresponding virtual environment in Prescan,and design the control strategy of “adaptive cubature Kalman filtering algorithm+MPC”.The simulation analysis and verification were carried out by "Carsim+Simulink+Prescan" co-simulation platform.The results show that the designed control strategy has good tracking characteristics for variable speed and large curvature conditions.Aiming at the strong nonlinear control problem of intelligent vehicles,the improved cubature Kalman filter algorithm has a higher precision of intelligent vehicle state parameter estimation.Combined with the model predictive control strategy studied,the trajectory tracking effect of vehicles at different speeds is effectively improved,which provides technical reference for enterprises to carry out vehicle running state parameter estimation and trajectory tracking strategy research and development.The project should have a certain value.