Calibration of a micro-traffic simulation model with respect to the spatial-temporal evolution of expressway on-ramp bottlenecks. (June 2016)

• Record Type: Journal Article
• Authors: Song, Rui; Sun, Jian
• Is Part Of: Simulation. Volume 92:Number 6(2016:Jun.)
• Page Start: 535
• Page End: 546
• Publication Date: 2016-06
• Subjects: Urban expressway -- on-ramp bottlenecks -- micro-traffic simulation -- parameter calibration -- transferability test; Computer simulation -- Periodicals; 003.3
• DOI: 10.1177/0037549716645197
• Languages: English
• ISSNs: 0037-5497

Abstract

On-ramp bottlenecks are usually the direct cause of traffic breakdowns on expressways, affecting the overall traffic operation of the expressway. A high-precision simulation model is the basis of an on-ramp bottleneck active traffic management strategy design and evaluation. This paper presents a practical and effective calibration method with respect to the congestion spatial-temporal evolution and two-capacity phenomenon of bottlenecks. An integrated criterion is proposed to evaluate how well the micro-simulation model can replicate the breakdown process. The widely used micro-traffic simulation model VISSIM is selected to simulate two on-ramp bottlenecks on an expressway in Shanghai, China. Firstly, 28 parameters of car-following and lane-changing behaviors are analyzed, and nine significant parameters of the operation at on-ramp bottlenecks are selected using sensitive analysis. Then, an orthogonal experiment is conducted to seek the best parameter combination. Finally, the calibrated simulation model is validated using a dataset of other days. Moreover, the transferability of the calibrated model is tested at Wuning bottleneck. The results show that the calibrated model can not only replicate the congestion evolution process and the special early onset breakdown phenomenon, but it can also apply to other similar on-ramp bottlenecks directly. The proposed calibration procedure, the integrated criterion for precision evaluation, and the selected significant parameters are meaningful to inspire rapid and effective calibration for other simulators.