Research And Implementation Of Autonomous Vehicle Tracking Method Based On Differential GPS

Autonomous tracking is not only the basis for unmanned driving but also the core for intelligent vehicle technology.GPS technology is commonly used in unmanned systems.Furthermore,differential GPS has higher positioning accuracy and lower cost in autonomous tracking.The core content of autonomous vehicle tracking includes three aspects: synchronous acquisition and storage of differential GPS and vehicle status information;fusion modeling and formation of target trajectories with collected differential GPS information and vehicle driving data;autonomous tracking algorithm to realize autonomous driving of driverless vehicles based on target trajectory.The main contents are as follows:(1)Using the BD930 receiver as a reference station receiver,the BD982 receiver as a rover receiver to collect positioning and heading signals for intelligent vehicles.This study traverses the number of delimiters in the associated instruction to extract and store the desired position and pose information.It analyzes the CAN message of the vehicle and then establishes two threads and design related algorithms to match and store GPS data and vehicle status information.Consequently,synchronized storage table of GPS data and vehicle status information is obtained.It ensures the synchronization of GPS and vehicle status information.(2)Converting the coordinates of the positioning data in BD982 receiver with Gaussian projection and modeling the converted data.In order to satisfy the continuity of GPS positioning data,this paper adopts B-spline fitting method to get the vehicle running trajectory curve based on the classic methods such as the least square method,Bezier curve method and B-spline fitting method.Aiming at the problem that the interpolation points on the trajectory are not associated with the vehicle status information,this study accurately predicts the state values of each interpolation point on the trajectory curve by Kalman filter method,so as to pledge the data fusion of vehicle running state and GPS curve.(3)Proposing a new prediction point model and a loop iteration method for solving the preview points based on the preview tracking model.The model improves the effectiveness of the preview point selection.In this study,an improved pre-tracking model the vehicle steering control and speed control methods in unmanned conditions are applied to realize the autonomous tracking of the vehicle.(4)Setting up a real vehicle experiment platform and carrying out an autonomous vehicle experiment.Compared the running trajectory and target trajectory of the vehicle during autonomous tracking,it validates the effectiveness of the autonomous tracking algorithm proposed in this paperTo conclude in fact,the research results of this paper have been applied and verified in the 2017 China Intelligent Vehicle Championship(CIVC).It lays a solid foundation for the research of the National Natural Science Foundation of China,which is the basis of this article.