Study and simulating mix traffic using SiMTraM

摘要

A Mix or Heterogeneous traffic consists of vehicles which have different physical and operating characteristics, even our country India falls in the criterion of mix traffic. There is also a variation of lateral and longitudinal gap between vehicles when taken into account in a single time frame in a selected stretch of study area. Homogeneous design methods and most of the simulation software's do not fit the heterogeneous situation, especially in non-lane-based roadways that populate the developing world. The purpose of the project is to study the traffic mix in some cities of India and then to find out the proportion of vehicles which finally forms the traffic. Video graphic surveys were laid out in Guwahati and Agartala using HD camcorders from footbridge and rooftops. The footages after converting into designated format were than feed into Trazer an image processing software which gave data such as vehicle counts, flow etc. The data generated from Trazer were analysed and further feed into SiMtraM to get the variance between the real and simulated data. The best mix by simulation giving inputs of various probabilities of mix of vehicles, even a result showing the difference of vehicle characteristic between the real world and simulation were found out. The validation of the software results with the real world data was statistically verified by T Test. In simulation different proportion of vehicle types was given as input so as to find out the effect of heterogeneity in the traffic. From the analysis it is seen that if the proportion of two wheeler is more than the flow and speed is more. Different car following techniques such as Intelligent driving model, Krauss model, Wiedemann model were also tested and the krauss model was found to be better.