Dynamic Model Of Vehicle With Five-DOF And Adaptive Model Predictive Control

The increasing number of automobiles in China will bring a series of traffic problems,such as road congestion,accident rate rising,etc.Therefore,the intelligent transportation system with low accident rate and high road traffic efficiency has been put forward,which has been studied deeply by many scholars.As an important part of ITS,the key of driverless vehicle is to establish accurate vehicle control dynamic model and design reasonable controller.Vehicle Control System is a multi-variable nonlinear system with multi-constraints,and MPC is good at dealing with multi-constraints.Therefore,on the basis of the five-degree-of-freedom vehicle handling dynamics model,MPC is applied to the control of unmanned vehicle and adaptive control is introduced.This paper is organized as follows:(1)In order to describe the actual movement of the vehicle in the complex road conditions,this paper uses the Lagrange Equation method on the basis of the vehicle tire model formula,based on the lateral,Yaw,roll,pitch and Steering Angle,a five-degree-of-freedom vehicle handling dynamic model is established.Firstly,the variables in Lagrange Equation are derived in detail,and the total kinetic energy includes the kinetic energy of the suspension,the body and the steering system,the body potential energy is obtained from the combination of the roll and pitch potential energy,and the steering potential energy is taken into account The generalized forces take into account the lateral,Yaw,roll,pitch and steering forces.Then the total kinetic energy,potential energy,dissipative energy and the generalized forces of the vehicle are substituted into the Lagrange Equation,and a five-degree-of-freedom dynamic model of the vehicle is established.(2)Because MPC has poor rod performance under complex road conditions,the AMPC is designed in this paper.Firstly,a linearization method based on the reference system is applied to the vehicle five-degree-of-freedom model to obtain the continuous state equation.Secondly,the first-order difference quotient method is used to discretize the continuous equation of state,and a nonlinear five-degree-of-freedom dynamic model linearized at any reference point is obtained.Finally,the MPC is improved,and the stiffness parameters of the vehicle are identified by the recursive least squares,and the AMPC is designed.(3)Simulation research.Aiming at the vehicle five-degree-of-freedom dynamics model,the vehicle parameters are introduced,and the simulation results with Matlab/Simulink/Car Sim show that the dynamics model is accurate.Aiming at the adaptive MPC,under S-Turn and double-shift working conditions,the joint simulation is carried out by using Matlab/Car Sim.Through the Comparative Study of the Control Effect of the AMPC and MPC on the driverless vehicle,it is concluded that the AMPC is better than MPC in the path tracking effect,the conclusion that vehicles run more stably.