Short-Term Traffic Flow Forecasting for Improved d Estimates of Freeway Incident Induced Delays

摘要

Freeway traffic incidents not only threaten drivers’ safety but also cause severecongestion, especially in urban areas. Congestion induced by incidents results in extra travel timeand travel cost. Quantifying incident induced delay (IID) would help people better understandthe real cost of incidents, maximize the benefit-cost-ratio of investment on incident remedyactions, and develop active traffic management and integrated corridor management strategies.In this study, a new approach for IID quantification was developed. Combining a modifiedqueuing diagram and short-term traffic flow forecasting techniques, the new approach can beeasily applied to measurements from existing loop detectors and overcame the limitation broughtby the zero vehicle-length assumption in traditional deterministic queuing theory. Furthermore,this approach is easy to apply because it uses only the volume data from loop detectors tocompute IID. Verifications using the video-extracted ground truth IID data found that the IIDestimation errors by the new approach were within 6% for the two incident cases studied in thisresearch. This implies that the new approach is capable of producing fairly accurate freeway IIDestimates using volumes measured by existing traffic sensors. This approach has beenimplemented in a web-based system, which enables quick, convenient, and reliable freeway IIDestimation in the Puget Sound region.