Research On Advanced Driver Assistance Systems In Freeway Environment

Autonomous vehicle technology has great significance for people to traffic safely and build smart cities.Due to the reliability of the sensor and the accuracy of the control algorithm,the current autonomous vehicle technology is not very safe and reliable.Prior to the maturity of autonomous vehicle technology,driving assistance technology was commercialized and applied.This paper focuses on the Advanced Driver Assistance Systems(ADAS)in freeway environment,obtains traffic scene information based on 3D lidar technology and designs driving behaviors decision including adaptive cruise control,automatic emergency braking and active lane change assistance.In addition,the autonomous vehicle motion planning control system is established,and a variety of simulation conditions are set up to complete the software-in-the-loop test analysis.Firstly,the paper establishes an environment-aware system based on 3D lidar technology.The ground point clouds are filtered out by the segmentation method of based on ray slope threshold,and the road passable area is segmented.The ground segmentation result is compared with the result of the plane model segmentation method of based on random sample consensus algorithm.For the problem of uneven distribution of point cloud density,divide the multi-threshold clustering area,based on Euclidean distance,the dynamic obstacles are clustered.After completing clustering,the interactive multi-model probability data matching algorithm is used to match the obstacle information.The unscented Kalman filter technique is used as the underlying obstacle tracking algorithm.The effectiveness of the algorithm is verified by the real lidar point cloud data.The ground segmentation can be completed in real time,the dynamic obstacle target can be continuously tracked,and the target motion state information is accurately obtained.Then,based on the obstacle information acquired by the sensing system,combined with the current state information of the autonomous vehicle,the driving behavior decision system based on the finite state machine is designed,including six system states capable of realizing three kinds of assisted driving functions,and the vehicle expected speed is adjusted by outputting the acceleration signal.The vehicle dynamics model is established under the premise of multiple conditions.The path planner and path tracking controller are designed based on the model predictive control algorithm,and the local expected path is reasonably planned in combination with the behavior decision system.Finally,the Pre Scan/Simulink co-simulation platform is established to carry out the software-in-the-loop test.Considering the actual traffic environment designs multiple simulation test conditions,simulation test results are analyzed,including adaptive cruise control,automatic emergency braking and active lane change assistance driving functions.The results show that the behavior decision algorithm based on the finite state machine can reasonably convert the driving state according to the environmental changes,realize the follow-up of the target vehicle speed,and effectively take emergency braking in case of collision danger.The system can complete motion planning by model predictive control algorithm,and control the autonomous vehicle to smoothly track the re-planned local expected path to complete the lane change overtaking action,as a result,the requirements of the advanced driver assistance system can be met.