Local Path Planning For Autonomous Vehicles In Urban Road Environment

With the increasing demand for vehicle safety and comfort,the research on the autonomous vehicles is getting more and more deeper.In the process of local path planning,the surrounding environment information is obtained at first,then the next movement is decided through the analysis of the surrounding environment information.The research on lane lines identification,forward vehicles identification,lane change decision-making,lane change trajectory planning and dynamics analysis is carried out.Main work contents are as follows:(1)In view of the urban road surface structural characteristics,a lane lines identification method based on Hough transform is proposed.The original road images are pretreated firstly,and the lane lines are extracted by Hough transform,through considering the shape features of structured lane lines.It is verified that this method can well identify the lane lines in the MATLAB environment.(2)A forward vehicles detection method based on laser radar and monocular vision is put forward.The radar raw signal is preprocessed firstly.Then the filtered data is processed by clustering.The ROI(region of interest)of vehicles detection is determined preliminarily and the filtered data is converted to the image coordinate.The symmetry measurement of normalized entropy of the preprocessed images is obtained to detect preceding vehicles.Compared with other vehicle detection methods,it is verified that this method can well identify the vehicles ahead and the time consuming is short.(3)A preliminary study on the lane change decision-making of autonomous vehicle is carried out by analyzing the specific requirements of the driving task of autonomous vehicle.Lane changing safety distance model and following safety distance model of autonomous vehicle are established.According to the requirements of the lane change trajectory,The quintic polynomial model is used to plan the lane change trajectory.The optimization objective function is designed by considering the comfort and stability in the lane changing process,and the sequential quadratic programming(SQP)optimization algorithm is used to optimize the lane change trajectory,the lane change trajectory of different initial speed is solved.(4)The vehicle dynamics model is established in Adams/Car,the corresponding simulation environment is set according to the trajectory planning results,the trajectory of different initial speed is tracked.The results show that the trajectory planning results conforms to the vehicle dynamics requirements,the reasonableness of the lane changing trajectory is verified.