Research On Path Tracking Model Predictive Control Methods For Autonomous Driving Vehicles

With the continuous progress of science and technology,the automobile industry is also undergoing profound changes.The design and manufacturing concept of modern automobile is committed to building it into an advanced autonomous intelligent automobile product.This has directly promoted the development of autonomous vehicles,and has also prompted autonomous driving technology to become an important direction for the development of modern automobile.In the research of autonomous driving technology,the trajectory tracking problem of the planned path of autonomous driving vehicle is an important topic that needs to be solved and optimized.Based on the Model Predictive Control(MPC)theory,this paper studies the trajectory tracking control strategy for the planned path of autonomous driving vehicles.The main research content of this article has the following three points:(1)In order to study the trajectory tracking of autonomous vehicle on the planned path.The traditional MPC theory is introduced to design the trajectory tracking control strategy for the planned path of the autonomous driving vehicle.First of all,according to Newton’s laws of mechanics,the kinematics mechanism of the vehicle is modeled.According to the MPC theory,the mechanism model is used as the system predictive model.Secondly,this part designs control input constraints and output constraints,and it uses quadratic programming numerical simulation to perform online rolling optimization to solve the objective function.The optimal control law obtained by the solution is applied to the controlled vehicle platform to realize the trajectory tracking of the planned path by the automatic driving vehicle.Finally,through a simulation experiment,the control strategy designed in this research is verified to be effective when the autonomous vehicle tracks the planned path,and the goal of the autonomous vehicle’s trajectory tracking of the planned path is initially achieved.(2)Aiming at the problems of poor tracking stability and weak adaptability to target speed changes during trajectory tracking,a MPC strategy based on terminal state equation constraints is further proposed.First of all,the dynamic modeling and analysis of the vehicle are carried out.The system prediction model is established based on the dynamic model of the vehicle,as well as the state variables that characterize the system are established at the same time.Secondly,this part has designed terminal state equation constraints and other constraints.Subject to the design constraints,the objective function is optimized online,and the obtained optimal control law is applied to the controlled vehicle platform.Finally,it is proved that the designed closedloop control system is stable,and simulation experiments are carried out to verify that MPC trajectory tracking control strategy based on the terminal state equation constraint is effective,and its following stability has been significantly improved.(3)In the research of the whole subject,in order to further improve the vehicle’s following speed and stability,so that the following error can quickly converge.A trajectory tracking MPC strategy based on Predictive Input and Output’s Contractive Constraint(PIOCC)is proposed.First,comprehensively considering the previous research on vehicle kinematics modeling and dynamics modeling,vehicle kinematics modeling is used as a predictive model in this part of the research.Secondly,in order to expand the scope of the feasible solution of the objective function,soft constraints are introduced.In addition,to avoid the divergence of the closedloop control system caused by the short prediction time domain,the system input and output contraction constraints are designed in the prediction time domain,and the vehicle trajectory tracking problem is transformed into the solution of the optimal problem is solved by online quadratic programming,and the first element of the obtained optimal solution sequence is applied to the controlled vehicle platform,and the process is cyclically optimized.Finally,it is proved that the predictive control system of the trajectory tracking model based on PIOCC is asymptotically stable.It is verified by simulation examples.The designed control strategy has a good effect on the planned path trajectory tracking of autonomous vehicle,The following speed and stability of the vehicle have been greatly improved,and the following error can quickly converge.