Research On Road Identification And Suspension Control Of Driverless Vehicles

At present,smart cars have become one of the hot spots in the world.It can perceive the changes of vehicle in advance through the machine vision system on the body and transmit it to the various control systems,which greatly improve the response time of the control system.Because single recognition system may be inevitably lead to misunderstandings,the perception method of multi-system fusion become a hot spot for future research.The thesis relies on the National Natural Science Foundation of China Youth Fund Project "Research on Adaptive Switching of Vehicle Active Suspension Damping Control under Intelligent Road Sensing"(51605213),the binocular camera and radar system are used to recognize the rectangular bump in front of the vehicle,and its features are extracted through specific image and point cloud processing for establishing the suspension control.Then the active suspension LQG controller is designed and comprehensive evaluation system of car ride comfort is put up to analyze ride comfort.The main research contents are as follows:Firstly,the information acquisition system composed of binocular cameras and radars are installed on the driverless vehicle,the rectangular bumps are set up in front of the vehicle,and a large number of images and point cloud are analyzed in the computer to product a database for extracting feature information of rectangular bumps.Secondly,the images are grayed out,marginalized,extracted the region of interest,the pixel ratio algorithm is used to extract the height of the rectangular bump,at the same time,the rectangular bump point cloud and the ground point cloud are segmented.The TOF ranging algorithm is used to extract distance information,the extracted feature information is used to built the corresponding virtual road input model in Matlab/Simulink.Thirdly,the active and passive suspension system model of vehicle are built up,the active suspension LQG controller is designed.The vertical acceleration of vehicle body,the suspension dynamic stroke and the tire dynamic displacement is selected as the control targets,and the analytic hierarchy process is used to determine the weighting coefficient for the LQG controller.Before the vehicle reaches the rectangular bumps,the optimal feedback force and the application time of the feedback force are calculated.Finally,the comprehensive evaluation system for vehicle ride comfort is constructed based on the ISO2631 evaluation standard.The basic evaluation method is used to calculate the weighted acceleration root mean square value for random road input,and the auxiliary evaluation method is used to calculate the vibration dose value for pulse road input.In addition,the vibration decay time is also calculated according to the time-domain curve,and the comprehensive evaluation value is obtained to evaluate the ride comfort of driverless vehicle.The research results show that the use of binocular cameras and radar recognition system can help to provide parameter information for active suspension control in advance.The active suspension using LQG controller can greatly improve the passenger’s ride comfort,and the comprehensive evaluation system constructed can effectively evaluate car ride comfort.