Research On Automatic Driving Technology Based On Laser Scanning Radar

Automatic driving technology contains many technologies such as environment perception,planning and decision making and automatic control which is a hot research topic at now.Laser scanning radar has the advantages of strong anti-interference ability,high detection precision,long detection distance and so on.It is the main sensor for environmental perception.In the process of vehicle driving,it is based on GPS positioning,through the path decision planning and the perception of real-time environmental obstacles,to achieve automatic vehicle driving,is the current mainstream technology solutions.Based on the 16-line laser scanning radar,this topic studies the environmental perception technology in the road where the car travels and combines the environmental perception results with GPS data to study the planning and decision-making technology for the car to drive automatically.From the aspect of laser scanning radar's environmental perception,the traditional two-dimensional grid map can't obtain the contour features of the obstacle section.This topic constructs a three-dimensional grid space based on the real scene,pre-processes the original point cloud information output by the laser scanning radar,transforms the spatial coordinates and projects it into the set three-dimensional grid space.On this basis,the road environment is perceived.Based on the structured road model,the topic extracts road edge points and uses the least square method to fit road edge points to realize road edge detection.The vehicle pays more importance to the obstacles on the inner side of the road during driving.Based on the edge detection of the road,the subject removes the obstacles on the outer side of the road,clusters the obstacles on the inner side of the road through the improved DBSCAN algorithm,classifies and identifies the obstacles to be measured according to the geometric characteristics.Finally,the position information and category of the obstacles relative to the vehicle are obtained.The method has high accuracy and strong real-time performance.In accordance with the limitation of using single sensor scene,this topic combines the road environment sensing results of laser scanning radar and GPS pose information to study the planning and decision-making technology of car automatic driving.The automatic driving car ensures the vehicle to run according to the preset track macro navigation by processing and analyzing the position and posture information of the GPS.According to the road environment perception result of the laser scanning radar,the subject uses a proportional differential control algorithm based on angle deviation to control the vehicle's macro-navigation driving and then designs a fuzzy controller to control the car to drive automatically according to the driving behavior rules.Besides,the subject has set up an automatic driving experiment platform,studied the vehicle speed and steering control algorithm,then designed the system software using the software design idea of " high cohesion and low coupling",tested and verified the automatic driving technology researched in the subject on an automatic driving car.The results of the study are applicable to the occasions where the speed of vehicles is not high and the road conditions are simple,such as campus express delivery,campus tour guide and factory cargo handling.