The FRESIM simulation model has been calibrated for an expressway in Singapore. This paper presents the application of genetic algorithm as a search method for finding a suitable combination of parameter values. The calibration is based on field data collected in a typical evening peak period over a 5.9 km segment of Ayer Raah Expressway. The calibrated parameters, including free-flow speed and minimum distance and sensitivity factors in car-following, enable FRESIM to produce 30-second loop detector volume and speed output (averaged across all lanes) closely match with the field data. This is a significant improvement form using the default values.
展开▼