ASSESSMENT OF CENTRALITY PROPERTIES OF AKURE ROAD NETWORK

Author(s)

Seun Daniel Oluwajana - Federal University of Technology, Department of Civil Engineering, Akure, Nigeria -

Olufikayo Oluwaseun Aderinlewo - Federal University of Technology, Department of Civil Engineering, Akure, Nigeria -

Adebayo Oladipo Owolabi - Federal University of Technology, Department of Civil Engineering, Akure, Nigeria -

Silvana Vivian Croope - Delaware Department of Transportation, Delaware, USA -

Abstract

Three 1-square mile of different urban street networks were extracted from Akure road network and developed into primal graphs. The resulting edges and nodes in the network were described by their distance matrix. Shortest paths across the nodes were calculated as well as the centrality measures. The edge length distribution analysis of the network studied showed a single peak distribution found in cities with a pattern realized over a historical process, which was out of the control of any central planning agency. Further investigation into the correlation between centrality measures showed that they captured different properties with correspondingly low values. Degree centrality was found to be correlated over the network. This was linked to the spatial restraint imposed on the network by land use. Information centrality values revealed that not all nodes in a network perform the same function since the network was found to perform better when some nodes were deactivated. However, analysis carried out revealed that the betweeness centrality and information centrality values, which capture the load and the ability of a network to respond to deactivation of a network, were of utmost importance in assessing the network.