Development of Mixed Traffic Microsimulation Model Calibration for Signalized Intersections

Traffic microsimulation modeling is growing in demand due to its easy applicability and its efficiency in checking the effectiveness of any traffic management system prior to its application. In developing countries, the application of traffic management systems becomes complex due to the heterogeneous nature of traffic in these countries. Hence it becomes necessary to calibrate the microsimulation platform before applying any traffic management system to it. In this study, an approach for calibrating VISSIM for heterogeneous traffic conditions is developed. The methodology was applied to two intersections in Ahmedabad city. The sensitive driving behavior parameters are identified by carrying out multiple sensitivity analyses by considering capacity as a measure of effectiveness (MOE). In order to check the effect of group parameters on MOE, a two-way Analysis of Variance ( ANOVA) method was used. The study identified eight sensitive parameters affecting MOE. Based on these eight parameters, two equations with respect to the intersections were developed using linear regression. The optimum value of the parameters is obtained by the Genetic Algorithm by keeping 95% confidence level. Validation is done by calculating GEH statistics and RMPSE of the traffic volume.