Research On The Vehicle Collision Warning Method Based On Trajectory Prediction In The Internet Of Vehicles

The traffic intersection is the core hub of the transportation network,but the traffic intersection has become one of the areas with high traffic accidents due to its many branch intersections and complicated traffic conditions.To achieve effective collision warning in intersection areas is one of the most important issues in the traffic safety.In this paper,we propose a Clothoid curve based Intersection Collision Warning scheme,called CICW,in Vehicular Ad Hoc Networks.In CICW,we first propose a vehicle trajectory prediction model based on straight-line driving and a vehicle turning trajectory prediction model based on cubic convolution curve.Through this model,the vehicle can establish a trajectory prediction equation.By utilizing the vehicle’s own state information,electronic maps,DGPS data,and neighboring vehicle state information from periodic beacons,each vehicle can predict the intersection of its own trajectory and the predicted trajectory of an adjacent vehicle based on the solved equation.In this paper,the vehicle can also fully consider the size information of the vehicle through the local scene reduction method,find the earliest point where the collision may occur from the potential intersection point of the driving track,and reach the time difference of the earliest point of the collision according to the corresponding possible collision.Whether it is less than the set threshold to issue an alert.For the case where there is an error in the DGPS positioning accuracy,the corresponding scheme is also proposed to be corrected in this paper.At the same time,in order to predict the trajectory of the vehicle more accurately,this paper also adds the discussion of the key points of the traffic intersection.By studying the characteristics of the vehicle trajectory,the key points are predicted in advance to ensure the accurate trajectory model of the established vehicle.By extensive simulation on JAVA platform,we evaluate the performance of the proposed scheme.The results show that our scheme achieve higher collision warning accuracy and lower error warning ratio compared with the existing schemes.