Research On Dynamics Modeling And Empc-based Lateral Trajectory Tracking Control Of Intelligent Vehicle

In recent years,the world's major automobile manufacturers and major research institutions have invested heavily in the development and research of intelligent vehicles,which has become the development trend of the future automobile industry.As a core issue of intelligent vehicle development,lateral motion control has attracted wide attention.One of the main problems for an intelligent vehicle is to address its real-time ability to track the transverse trajectory under multi-constraints.Based on model predictive control(MPC)methodology,this paper proposes a novel algorithm design based on explicit MPC for trajectory tracking.It utilizes the MPT toolbox,and provides an offline fast solution.Based on the simulation on Simulink/Carsim simulator,it shows that the offline solving method has the same tracking performance compared with the online solving problem and the computation speed is significantly increased.This paper has provided a new approach for improving the computation speed of trajectory tracking problem in case of strong nonlinearity and multiple constraints.As an important part of intelligent vehicle project of Brilliance Auto Group,the developed controller is installed on Brilliance V3 sport-SUV as a test platform,and the site test are carried out.First of all,In order to cope with the problems encountered in current controller design that state constraint,input and output constraint and delay cannot be addressed effectively,this paper presents a MPC control for trajectory tracking and formulats the optimization problem of the trajectory tracking based on the kinematic and dynamic vehicle model.The cost function weight coefficient and steering mechanism saturation constraint of optimization problem are defined for the coordination betwe tracking performance and stabilitySecondly,After transforming the MPC problem with constraints to a Quadratic Programming(QP)problem,The simulations were performed at different speeds to test both the performance of trajectory tracking and the lateral stability of the vehicle based on the MPC method by a high fidelity vehicle model simulator,Carsim.It proposed the explicit MPC(EMPC)method to solving the problem off line to improve the computation speed.Based on the simulation on Simulink/Carsim simulator,it shows that the offline solving method has the same tracking performance compared with the online solving problem by YALMIP toolbox.The computation speed is significantly increased.This paper has provided a new approach for improving the computation speed of trajectory tracking problem with strong nonlinearity and multi-constraints.Finally,the Brilliance V3 test vehicle installed EMPC-based steering controller was tested for the control algorithm at the Brilliance Automotive Engineering Research Institute.The controller communicate with the test vehicle by UDP broadcast and CAN bus protocol,obtain vehicle status parameters and transmits steering control commands calculated by the controller.Tests have provided the trajectory tracking performance and the stability of the lateral controller designed based on EMPC.