Motorized Level of Service Classification for Urban Uncontrolled Intersections

摘要: Service and total delay are considered for classifying MLOS intersections, respectively. GPS is used to collect travel time speed data turning movements that transformed into average values. Thirteen junctions from eight different cities in India form the dataset. Divisive followed by agglomerative clustering (DAC-HAC) algorithm applied as a two-step process obtaining service ranges. Validation of clusters performed based on Davies-Bouldin score, Calinski-Harabasz index, Silhouette gaps. Based DAC-HAC, uncontrolled intersections classified six categories (Cat-I, II, III, IV, V, VI). Results indicate classes “D”, “E” “F” have significantly higher ranges compared Highway capacity manual “control” indicating mixed traffic conditions. Most under fall Cat-IV, VI, having (> 60 s/ vehicle/ approach). Finally, validation results done geometric roadside environmental features.