Research On The Vehicle Position Prediction Based On The Vehicle Wireless Terminal Data

This thesis mainly studies the vehicle networked system,based on the user’s historical driving trajectory,constructs the position prediction probability model,and then combines the user’s current driving trajectory to predict the user’s next driving direction,predicts the user’s driving trajectory in advance,and then actively provides the user with the trajectory.Related traffic information services.The vehicle networking system mainly consists of three parts:the vehicle end,the cloud and the mobile end.The vehicle end collects the location information of the current vehicle and nearby vehicles through the GPS chip and the Lo Ra network,and uploads the data packet to the cloud through the GPRS wireless communication technology.The data parsing module in the cloud processes and stores the location data,and the historical trajectory analysis module performs data mining on the historical trajectory points of the vehicle,finds the motion characteristics of the user’s driving,and establishes a position prediction probability model.According to the probability model,the vehicle travel route is predicted.The mobile terminal is designed for the driver and the background administrator.The driver can monitor the position information of the vehicle in real time through the mobile terminal.The main work of this paper is the research of location prediction algorithm based on historical trajectory.The work includes:1.The design and development of the vehicle network cloud system mainly includes data analysis module,user center module,vehicle management module,vehicle position prediction module and data storage service;2.The location prediction algorithm research adds the abnormal position culling and insertion operation to the existing position prediction probability model algorithm,and improves the accuracy of key point extraction by repairing the abnormal position in the historical trajectory.At the same time,the extraction of key bifurcation points is improved.Conditions,the new shortest distance algorithm is used to judge,reduce the false positive rate of key points,and further improve the accuracy of position prediction.Finally,the position prediction effects of the two algorithms are tested by Python simulation.The results show that the improved position prediction algorithm has a good predictive effect on most driving trajectories and has high practical value.