Forecasting approach in VANET based on vehicle kinematics for road safety

Abstract

This paper deals with the forecasting of collision events for road safety. Using significant parameters of each vehicle, such as position, speed and direction, it is possible to contribute to improving the road safety. We present a collaborative forecasting module in intersection scenario for collision avoidance. The proposed module is focused on the estimation of these parameters using a kinematic model of each vehicle to generate the trajectories estimation. The first simulation results show and assess that the vehicle trajectories estimated with the suggested kinematic modelling are realistic in all critical cases. The main goal of the suggested forecasting approach is to detect and avoid collision. On the basis of these trajectories estimation, the future occurrence of the collision event can be calculated, an alert must be generated and this will trigger the forecasting module in order to avoid collision. In addition, the second part of the simulation proves that the proposed forecast scenario is excellent for collision avoidance.