Affinity propagation clustering for quantification of level of service at uncontrolled median opening

Abstract

Level of service at uncontrolled median openings on multilane divided urban roads is not defined for heterogeneous traffic condition prevailing in India. The LOS of uncontrolled junctions is quantified by the delay experienced by the lower priority movement. Therefore, delay experienced by the lower priority movement i.e. U-turning vehicles at an uncontrolled median opening has been studied to quantify the delay ranges for different LOS categories. Recently developed, affinity propagation (AP) clustering algorithm was employed to achieve the solution of the classification problem. Five validation parameters are analyzed to get the optimal number of cluster number which is crucial for the classification of different level of service categories. Affinity Propagation clustering is used to get the delay ranges for 6 different LOS categories. The range is less than 4 second for best operating condition and is greater than 35 second for the worst operating condition.