Network Centrality Assessment (NCA) as an alternative approach to predict vehicular traffic volume: A case of Colombo, Sri Lanka.

Abstract

This paper presents a robust, dynamic and simplified method to predict vehicular traffic volume in urban and regional geographies based on Network Centrality Assessment (NCA). The case study was conducted in Colombo, Sri Lanka. Study employed three types of graphs and three kinds of analysis methods to compute network centrality referring to four centrality parameters; for identifying the predictability of vehicular traffic volume. Findings stress that, the road segments graph based on geo-metric analysis method and the natural roads graph based on topological analysis method is far better in predicting the vehicular traffic volume and it is more appropriate to consider the multiple influence from multiple centrality parameters predicting vehicle volumes rather than strict being into the single best parameter. Hence, study concludes that it is more appropriate to employ NCA considering the multiple influence from multiple centrality parameters based on geo-metric and topological analysis methods in predicting vehicle volumes.