Feasibility study for automatic calibration of transportation simulation models

摘要: This paper presents a high-tech solution to meet the challenges in calibrating transportation simulation models. Like any software, model calibration prior its application plays crucial role producing reliable results. However, professionals face difficulties performing daunting tasks of for each network design satisfy targeted traffic flow demand, especially during data collection and distillation. Our innovative approach utilizes sensor geography networking technology seamlessly collect about real world network, traffic, driver behavior. is then distilled as needed by mining before feeding model. The validated automatically instantaneously reflect avoid typographical errors often involved with human intervention, resulting more accurate model.We conduct feasibility study our vision automation. research flexes multidisciplinary expertise simulation, geosciences, sensing/networking, knowledge discovery. As proof concept, we implement prototype that demonstrates how convert collected state department into format taken CORSIM, popular A running example shows encouraging