Research On Vehicle Collision Warning Algorithm Based On Vehicle Self - Organizing Network

Improving vehicle driving safety is one of the most important applications of the Vehicular Ad-hoc Networks (VANETs) and for the traffic and transportation system, the traffic intersection is as important as joints for human beings. In the traffic intersection, different lanes’traffic volume come to the same area, different intersection shape exists, the traffic condition also could be very complicated, these all make the traffic intersection to be a traffic accident high-risk areas and most likely to be a traffic bottleneck area affecting the traffic volume. So improving traffic intersection traffic safety becomes a hot point.In order to reduce the accident rate at the traffic intersection, combining the characteristics of the vehicle at the traffic intersection, we propose a traffic intersection collision warning algorithm by predicting vehicle trajectory. According to the characteristic that the travel trajectory of the vehicle when turning at the traffic junction roundabout in line is Clothoid curve, we can make predictions on the running track of the vehicle. The RSU can calculate vehicle’s track curve equation after collecting the information about the vehicle moving in traffic intersection. Then the cross point between different trajectory can also be calculated, the possible collisions in advance can be pre-warned, suggested speed can be given to avoid traffic accidents.To evaluate the usability and accuracy of our traffic intersection collision warning algorithm by predicting vehicle trajectory, we simulated the traffic intersection scene using MATLAB software and processed the measured experimental data. Experimental results showed that the algorithm can accurately predict the trajectory of the vehicle at the next moment; the error is less than the size of the vehicle itself; the accident early warning accuracy rate of more than 90% with high accuracy.