Traffic Forecast Based on Statistical Data for Public Transport Optimization in Real Time

Abstract

The intellectualization of transportation system is a relevant task for solving traffic-related problems such as vehicle traffic management and modern transport system improvement. The purpose of this work was to design a competitive pathway for real transport system to optimize the route planning. Modeling of the transport system refers to the problem of finding the K-shortest sustainable pathway in a multimodal network. The solution to this problem was fulfilled by applying a hybrid algorithm of an ant colony based on a differential evolution approach to the pheromone renewal and the division of the colony into teams. Experimental results showed that an advanced ants colony algorithm is highly efficient even with quite a small population of the colony. The obtained results were compared with those of the conventional ant colony algorithm. Due to its high efficiency, the elaborated method is applicable in improving the quality of the entire road network, especially in congestions and traffic jams, taking into account real-time traffic information.