Collision probability computation based on vehicle to vehicle communication

In this paper, we propose a method to compute the collision probability of two cars at a traffic intersection. The cars´ position, speed and intention are used to predict their trajectories, then the possible collision location is calculated, and the collision probability is computed as well. An experimental platform including simulated traffic intersection, an opti-track motion capture system, and a vehicle-to-vehicle communication system is built to demonstrate this method. By using the driver´s intention information, the collision probability curves are generated, and the results are compared with those of the method without considering driver´s intention. Experiment results show that the computed collision probability of this method is closer to that in the actual situations. Hence, it can aid drivers to make correct decisions at intersections. This method can also be used in future autonomous cars for collision avoidance through vehicle to vehicle communication.