Based On Multi-sensors Intelligent Vehicle Environmental Perception System Research

With the rapid development of automobile industry,intelligent vehicle is becoming a hot spot for automobile research.Intelligent vehicle is mainly composed of environmental perception,strategy,automatic control,and path planning.Among them,the environmental perception system is the basis of intelligent driving,which works by collecting environmental information around the vehicle,such as lane lines,pedestrians and other vehicles.In this paper,the main research contents:(1)Firstly,based on multi-sensors technology,we established the intelligent sweeper experimental vehicle platform and designed the environmental perception system algorithm.Furthermore,installing camera,long-range and short-range millimeter wave radar,and ultrasonic radar sensors in the experimental vehicle to detect all of the objects nearby,as well as the front lane.(2)Secondly,we concentrated on collecting the raw data millimeter wave radar from the vehicle CAN bus and designing data preprocessing algorithms,including bubbling screening algorithm and time-domain filtering algorithm for the purpose of stability following and detection.Also,the time-domain filtering algorithm laid foundation for multi-sensors data fusion algorithm.(3)Thirdly,we designed frontal objects detection level sensors and central data fusion algorithm for structured roads.The fusion algorithm can improve the objects detection probability largely and physical information accuracy.Then,applying the kalman filter to the objects which are detected by fusion algorithm can make the frontal vehicle following and detection more accurate.Finally,test the environmental perception system of the experimental intelligent sweeper vehicle.We set pedestrians,vehicles or bicycles as frontal detecting objects.The testing results showed that the multi-sensors data fusion algorithm designed in this paper can identified tracked objects reliably.Letting the bicycle approach to the experimental vehicle,we found that the data preprocessing algorithm can stability tracked bicycle target.Thus,the algorithm realized basic environment perception system functions,providing the vehicle comprehensive environmental information for vehicle controller and ensuring the safety of vehicle driving.