Nauka ta progres transportu (Oct 2014)
INVESTIGATION OF SCALE-INVARIANT PROPERTY OF ORGANIZATION SYSTEM OF TRAIN TRAFFIC VOLUME BASED ON THE PERCOLATION THEORY
Abstract
Purpose. The work is devoted to the study the property of scaling invariance of the organization system of train traffic volume on Ukrainian railways. Methodology. To prove the real network origin of Trains Formation Plan (TFP) destination to the type of so-called scale-invariant networks it is proposed to generate scale-free networks with different dimensions, Barabási–Albert type with parameters that real networks of TFP destination has and to investigate their structure on survivability using the procedure of percolation nodes. Percolation process is proposed to be considered as a modified version of the spatial movement of cars on the network by increasing the number of railway stations, which have lost the ability to perform the basic function to pass cars on TFP destination in terms of adverse effects (an accident, overload). Findings. Comparative analysis of percolation at random and targeted destructive impact on network nodes has shown matching with the results of real network percolation of TFP destination, which proves the existence of self-similarity. Comparable figures in percolation were: percentage of remote stations in the network, in which the network fragmentation occurs, the average inverse path between network nodes, the diameter of the graph structure, the size meaning of the second largest cluster in the network from the steps of destruction. Originality. For the first time the hypothesis of the existence of scaling invariance properties of the graph TFP destinations on the railways of Ukraine, which can be attributed to a class of the graph scale-free networks was confirmed. Existing knowledge in the field theory of scale-free networks can be used to describe the survivability of system transportation on the railways of Ukraine. Practical value. Based on the identified properties of system directions of train traffic volumes, it is possible to create a mathematical model in the future that will predict the behavior of the transportation system with network structure. Properties analysis of system survivability of train traffic volumes will optimize the use of capital investments to increase network capacity by identifying the most critical lines and stations that systematically affect the efficiency of the network as a whole.
Keywords
