Clustering Traffic Flow Patterns by Fuzzy C-Means Method: Some Preliminary Findings

Abstract

In this paper, performance of fuzzy c-means clustering method in specifying flow patterns, which are reconstructed by a macroscopic flow model, is sought using microwave radar data on fundamental variables of traffic flow. Traffic flow is simulated by the cell transmission model adopting a two-phase triangular fundamental diagram. Flow dynamics specific to the selected freeway test stretch are used to determine prevailing traffic conditions. The performance of fuzzy c-means clustering is evaluated in two cases, with two assumptions. The procedure fuzzy clustering method follows is systematically dynamic that enables the clustering, and hence partitions, over the fundamental diagram specific to selected temporal resolution. It is seen that clustering simulation with dynamic pattern boundary assumption performs better for almost all the steps of data expansion when considered to simulation with the corresponding static case.