Modeling And Simulation Of Environment Awareness System For Intelligent Driving Vehicles

In recent years,more and more people pay attention to intelligent driving vehicles.As an important part of intelligent driving,the environment perception system of intelligent driving vehicle has become a research hotspot of many scholars at home and abroad.In order to improve the accuracy and robustness of the environment awareness system of intelligent vehicles,this paper studies the application of vision sensors and laser sensors in the environment of intelligent vehicles。The environment perception system of intelligent driving vehicle based on camera and the environment perception system of intelligent driving vehicle based on laser radar are constructed.In the environment perception system of intelligent driving vehicle based on camera,lane line detection and environmental target detection of intelligent driving vehicle are mainly carried out,such as vehicles and pedestrians in the environment of intelligent driving vehicle.In the environment sensing system of intelligent driving vehicle based on lidar,the obstacle information identified by lidar is dedistorted,the ground data is divided,and the obstacle clustering is analyzed by using Euclidean clustering.In view of the differences and advantages and disadvantages of each sensor in the sensing environment,the software system of lane line detection system based on camera sensor is designed and developed for simulation analysis,and the demonstration test of the two systems is carried out.Based on the current application research results of intelligent vehicles,this paper conducts an in-depth study.(1)The environment perception system of intelligent driving vehicle based on vision sensor and lidar sensor is constructed.According to the working principle of vision sensor and lidar sensor,the sensor is calibrated by combining their calibration methods.(2)Based on the driving characteristics of intelligent vehicles,the lane line detection system of intelligent driving vehicles is constructed.(3)According to the diversity of target categories of obstacles in the perceived environment,K-means algorithm is used to obtain multiple groups of prior boxes adapted to target data,and YOLO-V3 algorithm is used for target recognition.(4)Make use of the high quality characteristics of lidar sensor to dedistort the data collected by lidar.(5)Segmentation ground processing of point cloud data based on ROS operating system,and cluster analysis of lidar point cloud information.(6)The lane line detection system based on intelligent driving vehicle is designed and developed based on GUI(visual interface)in Matlab.(7)Build a real vehicle test platform to test the accuracy of the built system.