Calibration of visual angle car following model based on real site traffic data

Traffic simulation models have increasingly been used due to low cost, time saving and ability to examine possible solutions for traffic-related problems without interrupting the real life traffic conditions during testing alternative scenarios. Recently, visual angle car following model has been studied and examined by the author for use in developing a microsimulation car following model. The proposed model has been tested for different angular velocity thresholds against real published traffic data by developing a simple-one lane microsimulation program. The results show that when the suggested angular velocity threshold of about 0.003 rad/sec is applied, the model will become able to replicate real life traffic movements. Root Mean Square and Error Metric statistical tests have been used to compare different selected angular velocity thresholds. A non parametric, Kolmogorov-Smirnov test indicates that the difference between the observed and simulated data is significant when the angular velocity threshold is below a value of 0.002 rad/sec. The aim of developing this car following model is to be used at a later stage in developing a micro-simulation model to represent traffic behaviour at motorway merges and to test the objectiveness of using ramp metering strategies on motorways.