Path Tracking Control Research For Autonomous Vehicle

The main problems with increasing vehicle production are traffic congestion,energy consumption and automobile safety.The autonomous vehicle completely excludes the interference of human factors,so it can solve the problem of traffic safety greatly,and the stable and reliable path tracking system is a necessary condition for the realization of industrialization of autonomous vehicles,therefore path tracking control problem of autonomous vehicle is studied deeply,mainly completed the following work:(1)The whole system architecture of an experimental platform Lifan 620 is introduced.the working principle and technical parameters of the vehicle sensor are described in detail,and the whole chassis system of the vehicle is designed,the automatic control of each execution module is realized,which lays the hardware foundation for the subsequent research of real vehicle experiment.(2)A control strategy of vehicle longitudinal system is studied,which is designed with the idea of layering,and the speed control is controlled by PID.The throttle and brake control of the lower controller is analyzed and calculated by the reverse longitudinal model.In order to ensure the stability and comfort of the vehicle during acceleration and deceleration,the throttle / brake switching rules are designed to avoid frequent switching between the two in the actual driving process.The effectiveness and robustness of the longitudinal control system algorithm are verified by the joint simulation of PRESCAN and Simulink.(3)Based on the improved Pure Pursuit tracking algorithm,the path tracking lateral control system of simulated driver model is studied,the system consists of two parts: the preview lateral error generator and the target heading angle feedback control system.The preview lateral error is determined according to the kinematics relationship between the vehicle and the road,while the feedback control system adopts the PID controller which simulates the driver model.Through the joint simulation experiment,it is verified that the algorithm can track the target trajectory well under different working conditions,and has a high tracking accuracy,which lays a foundation for transplanting the algorithm to the real vehicle environment.(4)Under the Simulink automatic code generation system platform based on the model development,the control program of the whole vehicle underlayer execution system is written,under the Linux system,the upper computer software is written with C++ language and QT as the development platform,the function of each component of the software interface is explained in detail,and the coordinate system conversion software is developed based on Visual Studio2010 software platform.Finally,in order to verify the effectiveness of the path tracking algorithm in this paper,the road section including typical working conditions is selected to carry out the real vehicle experiment,the experimental results show that the proposed horizontal and longitudinal control system has good control effect and can meet the requirements of driverless vehicle path tracking control.