Simulation of ITS on the Irvine FOT Area Using "Paramics 1.5" Scalable Microscopic Traffic Simulator: Phase I: Model Calibration and Validation

In this research, a promising ITS-ready microscopic traffic simulator, Paramics1.5, is thoroughlyevaluated both subjectively and objectively in preparation for its use for modeling ITS in SouthernCalifornia. A general and comprehensive list of requirements (evaluation template) for amicroscopic traffic flow simulator to successfully model ITS was first compiled and presented. Theevaluation template covered various aspects of modern transportation network modeling, including:[1] supply/control aspects, [2] demand/behavior aspects, [3] environmental-related aspects, and [4]simulator-performance aspects. The template was applied to Paramics1.5, a new and potentiallysuccessful ITS-oriented micro-simulator. The evaluation revealed Paramics1.5’s capabilities andlimitations as an ITS modeling tool. It was concluded that Paramics1.5 is an excellent ‘shell’ or‘framework’ for a comprehensive and extensive transportation simulation laboratory. Paramics1.5offers two very important and unprecedented features: high performance and scalability. To ourknowledge, this is the most promising approach to handle realistic real word traffic networks underITS. Also, Paramics1.5 offers very plausible detailed modeling for many components of theenvisioned ‘ideal’ simulator. Nevertheless, several limitations were also identified and discussed.Suggestions for improvements were also presented. Improving such limitations requires both directmodifications in the software itself, and maturation of the Application Programming Interface(API), one of Paramics1.5’ most important capabilities. The subjective evaluation was also followedby an objective evaluation based calibration of key driving-behavior parameters underlying -amongother things- the car following and lane changing models of the simulator.Calibration wasconducted using real traffic data from the California Testbed, headquartered at the University ofCalifornia Irvine. The calibrated model performed well during validation on a freeway link. On afull network, the vehicle release mechanism showed some time-lag in releasing demand onto thenetwork. This is potentially due to stacking of vehicles in memory before adequate headways arefound on the road to release the vehicles. Although the problem itself is simple, its effects on theresults were notable.