Research On Path Tracking And Obstacle Avoidance Control System Of Autonomous Vehicle

With the rapid increase of the number of vehicles,traffic safety and traffic congestion has become an urgent problem to be solved.With the continuous promotion of innovative artificial intelligence technology,the critical technologies of autonomous driving are advancing by leaps and bounds.Self-driving vehicles can solve traffic safety problems by eliminating the interference of drivers,and the overall controllability of the network of self-driving vehicles can largely avoid traffic congestion.As stable and reliable path tracking control system and autonomous obstacle avoidance control system are the necessary conditions for the realization of autonomous driving,this paper conducts in-depth research on the problems of path tracking control and autonomous obstacle avoidance control of driverless vehicles.The main research contents include the following three aspects:(1)Design the software and hardware architecture of self-driving vehicles.Refit the traditional vehicle steering system,braking system and speed control system,and stm32 MCU was used to design the closed-loop controller of the base control system.According to the requirements of the autonomous vehicle,install and configure its base system sensors and environment awareness system sensors.Based on the ROS(Robot Operating System),the distributed design of the automatic driving software system is carried out.The upper and lower data communication protocol is designed based on RS232 to realize the control of the base system and the signal feedback.Finally,the software and hardware system of the designed platform is verified by experiments.(2)Study the path tracking control algorithm of self-driving vehicles.The kinematics model and geometry model of the vehicle were established,and the established model was built by Matlab.This paper studies the common geometric path tracking algorithm-Pure Pursuit,and analyzes the effect of foresight distance on tracking accuracy,the adaptability of foresight distance to defined path and vehicle speed.Based on the research of Pure Pursuit,a novel path tracking algorithm considering the path curvature is proposed.Firstly,the least squares curve fitting method was carried out on the path curve fit to calculate the path curvature,and then the maximum allowable lateral acceleration is used to limit the vehicle speed,Finally,the current speed and the lateral tracking deviation are used to adjust the foresight distance adaptively.Matlab are used to simulate the proposed algorithm and verify its feasibility and tracking accuracy.The algorithm code was written by c-plus-plus and the experiment was carried out on the experimental platform to verify its stability and applicability.(3)Research on local obstacle avoidance path planning algorithm for autonomous driving vehicles.The advantages and disadvantages of common local path planning algorithms are analyzed,including artificial potential field method,genetic algorithm and fuzzy logic control algorithm.Proposed a new local obstacle avoidance path planning algorithm based on local path isometric transformation.Establish the expandable object model to objects classification(obstacles and none-obstacles).Search the obstacle avoidance path iteratively until all obstacles degenerate into non-obstacles.Finally,a polynomial curve is used for lane change path planning,vehicle follows the lane change path to the obstacle avoidance path and realize the obstacle avoidance.