Application of genetic programming clustering in defining LOS criteria of urban street in Indian context

摘要: India is a highly populated country having second largest road network in the world. Owing to boastfully population, congestion growing rapidly on urban networks. The level of service (LOS) not substantially defined for heterogeneous traffic flow with different operational characteristics. Defining LOS essentially classification problem. application cluster analysis worthiest proficiency solve such problem which genetic programming (GP) clustering, an evolutionary algorithm used this study. Five validation parameters are utilized examine optimal number clusters. obtain categories street classes. After acquiring clusters, GP clustering implemented free speed (FFS) data get ranges Again, enforced average travel speeds segments specify categories. Speed study collected using Trimble GeoXT GPS receivers fitted mid-sized vehicles five major corridors comprising 100 Greater Mumbai region. Result shows that FFS classes and lower than mentioned Highway Capacity Manual (HCM 2000) account physical surrounding environmental Also, expressed terms percentage found be from HCM 2010.