This paper describes the calibration and validation of a micro-simulation model in the analysis of a large urban network. The Paramics micro-simulation model was applied to model traffic operations within the City of Niagara Falls, Ontario. The model is used to assess traffic operations and various traffic management initiatives including deployment of Intelligent Transportation Systems within the tourist area in the City. The analyzed network consists of freeways, arterial and collector roads with over 90 signalized intersections. The Paramics model incorporates a modeling procedure in which movements of vehicles through the network are governed by origin-destination matrices on the basis of various assignment techniques. Over 17 traffic zones were used to model points of traffic generation and destination including high traffic generators such as car parks in the tourist area. The calibration efforts focused on the PM peak hour and involved estimation of representative origin-destination matrices for the network, selection of appropriate assignment method and comparison of modeled and observed traffic volumes at the screen line, roadway segment and intersection levels using a modified Chi-Squared statistic test. The model validation was undertaken by comparison observed and modeled travel times and qualitative queue assessments along various routes in the network. The analysis shows that application a systematic approach that includes the above steps can result in a well calibrated and validated model that provides reliable results in practical applications. The paper also discusses the challenges and suitability using micro-simulation modeling in the analysis of large urban networks and notes that, focusing further calibration efforts on selected sub-areas and corridors would be more appropriate for such cases.
