On Traffic Flow Regimes and Transitions in Signalized Arterials

The paper presents the formulation and application of a data driven methodological framework to identify traffic flow regimes and transitions on signalized arterials from point measurements of volume and occupancy. The identification of regimes is conducted by a wavelet-based fuzzy clustering approach, while transitional conditions are studied using Bayesian networks. The results from this data-driven approach indicate the existence of four distinct traffic flow regimes; these regimes hold in arterials with different geometric and signalization characteristics. An analytical model is also developed based on kinematic wave traffic flow theory to determine the boundary conditions among traffic regimes. This model provides strong evidence that the presented statistical approach is in agreement with a simple and elegant analysis including traffic parameters that are observable and measurable.