Traffic Simulation as Highway Design Tool

This paper is a summary of the research conducted to develop, test, calibrate and apply the digital computer model with actual roadway and traffic data collected from nine field locations on rural primary roads in North Carolina. The model is used to simulate traffic flow over a specified section of roadway, generally 5 to 6 miles long, at different volume levels ranging between 175 and 650 vehicles per hour. The process of calibration consists of making logical changes in key input variables until the simulation throughput data statistically matches similar throughput data observed at field sites. The technique of multiple regression is used to develop relationships between such input variables and other roadway and traffic variables. The model is sensitive to changes in such input variables as reflected by the throughput statistics and that it can realistically simulate the two-lane two-way traffic flow to evaluate the traffic consequences of highway redesign concepts.