Obstacle Detection Technology For Autopilot System

Autopilot technology includes artificial intelligence,computer science,signal processing and automatic control technology.Autopilot platform should have the basic functions of environmental awareness,autonomous positioning and navigation,obstacle detection and avoidance.Obstacle detection and avoidance is a key technology to ensure that it can safely reach the target position.Therefore,it is very important to study the related technologies such as obstacle detection and avoidance.The main research contents of this paper include the following aspects:(1)The hardware and software platform of the robot platform is designed.The main components of the hardware platform are analyzed.The RPlidar A2 lidar used in this paper and its main technical parameters are emphatically analyzed.The communication mode between RPlidar A2 and PC is also introduced.At the same time,the top-level software framework of the robot platform is designed by using the Robot Operating System(ROS).(2)Comparing with several existing environment representation methods,grid map is chosen to describe the environment of the robot.The kinematics model of the differential mobile robot is established and its kinematics principle is analyzed.The SLAM problem is solved by mapping first and locating later.The static binary Bayesian filtering theory is analyzed and applied to occupied raster map construction algorithm to realize map construction.Experiments verify the effectiveness of the occupied raster map construction algorithm,and the resolution of the constructed raster map can reach 5 cm.(3)Several commonly used robot localization algorithms are analyzed and their advantages and disadvantages are compared.The mathematical principle and key technology of particle filter are studied.Emphasis is placed on the Monte Carlo localization algorithm.On this basis,the adaptive Monte Carlo localization(AMCL)algorithm is adopted to locate the robot,and the simulation is carried out to verify the function of the algorithm.The positioning effect of AMCL location algorithm is verified by experiments.(4)The common obstacle detection schemes are analyzed and compared.Laser radar is chosen to sense the environment around the robot.By means of mean filtering and nearest neighbor clustering processing of lidar data,obstacles in the environment can be detected and the number of obstacles can be found well.The feasibility of the algorithm is verified by experiments.VFH algorithm is used to realize obstacle avoidance of robots.The principle of obstacle avoidance of the algorithm is introduced,and the effectiveness of the algorithm is verified in real scenes.