Research On Path Planning And Tracking Control Of Autonomous Vehicle

In recent years,the rapid development of advanced technologies such as cloud computing and artificial intelligence has promoted the technological revolution in the information age.As the application of artificial intelligence technology in the automotive industry and transportation field,autonomous driving has received close attention from research institutions and even the national level in the world recently.Autonomous vehicles have Internet communication technology,which can sense the driving environment in advance and make safety warning,pointing out a new direction for improving road traffic safety.Path planning and path tracking are the key technologies of autonomous driving system.Ensuring the quality of planned path and tracking accuracy is an important prerequisite for the safe and stable running of autonomous vehicles.In this thesis,the improved artificial potential field method with simple structure and good real-time performance is selected and used in vehicle path planning module.An optimized model predictive tracking controller is designed to accurately track the target path.Specific research contents are as follows:Firstly,the vehicle three degree of freedom dynamics model and the magic formula tire model are established,and the mechanical properties of the tire model are studied.Under the premise of small angle hypothesis,the dynamic system model is simplified,and the model is approximately linearized.Finally,the discrete linearized state space time-varying model is obtained by the first-order difference quotient method,which lays a foundation for the model predictive controller in Chapter 4.Secondly,the traditional artificial potential field method in path planning is improved.By adding virtual target points,the problem of falling into local minimum value in the process of vehicle running is solved,and the problem that the target points cannot be reached in the process of vehicle path planning is solved by improving the repulsion function.In this thesis,the road boundary repulsion potential field is established to reflect the influence of the road boundary repulsion on the autonomous vehicle driving on the road.The velocity potential field of obstacles is added to the improved repulsion potential field to consider the influence of dynamic obstacles on the driving process of vehicles.Then,the optimal objective function of the model predictive controller is established based on the designed predictive model.By adding vehicle dynamics constraints and tire yaw angle constraints to the system,the stability of the vehicle during driving and the accuracy of the vehicle tracking the target path are ensured.Finally,in order to verify the tracking effect of the model predictive controller established in this chapter,a joint simulation platform composed of Car Sim and Matlab/Simulink is established,and a double line moving curve is set up for path tracking simulation experiment.The simulation results show that the state curve of the model predictive tracking controller is continuous and smooth when tracking the reference path of the double shift line,and the tracking performance of the controller is fine.Finally,the real-time planning and tracking control performance of the autonomuos vehicle in the actual driving process are verified.In this thesis,Car Sim is used to set up the vehicle model and road environment information,and the path planner and path tracking controller are established based on Simulink.The joint simulation experiment is carried out using Car Sim/Simulink platform to simulate the lane changing and overtaking scenes.The simulation results show that the real-time planning performance is good,the accuracy is high,and the tracking effect is fine.