Simulation of the Adaptive Multivariate Exploration for Routes Guidance

Abstract

Nowadays, Navigation systems have become very popular and necessary for travelers from all around the world. Navigation systems give travelers all the information they need and guide them quickly and comfortably to get to the destination. Pre-trip planning regarding road and traffic conditions can enhance driver's knowledge of the situation in road networks and can assist in drivers' decisions concerning routes and departure times. Many factors such as the route density, traffic volume, departure time, destination, workday, working time, holiday period, tournaments and events are important to find the best route under the criterion. In this paper presented two new models with the purpose of recommending travelers the best route to destination. These two models are the multivariate class exploration and route guidance models. The multivariate class exploration model analyzes the network-flow patterns in order to investigate some important factors such as demand conditions, compliance with information, speed and variance of travel-time and calculate a travel-time reduction. A route guidance model was used to find efficient routes (minimizing travel time) for traveling. This paper conducted experiments on a real-world dataset collected from the OpenStreetMap. The accuracy of the proposed model's predictions was determined. The results show that the predictions given by the models are accurate and can be used in real-life situations.