Some Guidelines for Selecting Microsimulation Models for Interchange Traffic Operational Analysis

摘要

There are several commercially available traffic simulation packages that are able to model the full range of interchange ramp terminals. Each of these packages has pros and cons for simulating various types of interchanges and traffic control plans. Previous research has studied specific simulators in terms of their capabilities for simulating single urban point interchanges (SPUIs) and diamond interchanges. There is no guidance provided however on how to generally select appropriate simulators based on their capabilities and internal algorithms for analyzing specific interchange design and control scenarios. This paper focuses on identifying the elements that should be available in a simulator in order to evaluate a specific interchange scenario (including type, geometry, and traffic control characteristics). The paper does not identify all the specific packages that are appropriate for a specific scenario because these evolve constantly; even though the current version of a package may not support a particular function, future versions of that package may incorporate it. Thus, identifying the most appropriate package would very quickly become obsolete. It is up to the analyst to examine whether a particular characteristic or algorithm is present in a specific package. To accomplish this, three simulators were selected and studied: AIMSUN, CORSIM, and VISSIM. Data from two interchanges (one SPUI and one diamond interchange) at Arizona were obtained and used in the assessment of these packages. The model parameters and assumptions are examined for each model. Simulation is conducted first by using default parameter values followed by a calibration process to adjust parameters related to driver behavior, vehicle performance, and others. In conclusion, several elements are identified as critical in simulating interchanges, which fall in the following categories: (1) the capability of representation of specific geometric characteristics; (2) the capability of simulating specific signal control plans; (3) calibration needs and accuracy in comparison to field conditions; (4) the extraction of specific performance measures from the simulator; and (5) other observations from the research.