Artifical intelligence based vehicular path prediction for inter-vehicular communication and its application

Road transportation has an indispensable role to humanity though it has lots of negation in terms of economic, environmental and social factors also especially the dangers on human life and health. One of the approaches which aim to make the road transportation more efficient and safer is Inter Vehicle Communication (IVC). This approach led to emergence of Vehicular Ad-Hoc Networks. The aim of IVC is to provide safer, more entertaining and more comfortable transportation in the way that vehicles share their own various data with other vehicles and the road infrastructures. The basis of the safety applications is that vehicle in move identifies the possible dangers and eliminates them by processing the data obtained from the vehicles and infrastructure in the environment and its own navigation and dynamic data. The most possible danger occurring on road transportation is vehicle collisions. Prediction of geographical position of a vehicle in move at a moment of the future is called as Vehicular Path Prediction. Two of the most important factors in the suggested path prediction models should be high accuracy and low computation time. In this study, a vehicular path prediction model which has high accuracy and fixed low computation time has been developed by using ANN for highway and rural roads. The development and the performance test processes of the model have been done as real time applications. A system is developed which predicts vehicle path by measuring geographic position and dynamic values of vehicles in move, shares the results with other vehicles via ad-hoc network and alerts the driver to the possible dangers.