Implementation and Evaluation of Weather Responsive Traffic Estimation and Prediction System.

摘要

The objective of the project is to develop a framework and procedures for implementing and evaluating weather-responsive traffic management (WRTM) strategies using Traffic Estimation and Prediction System (TrEPS) methodologies. In a previous FHWA-funded project, a methodology was developed and tested for incorporating weather impacts in off-line TrEPS. This capability is now included in on-line TrEPS, which interacts with multiple sources of local real-time information, to provide operators with predicted traffic states under the current and future weather conditions. The main goal is to support the decision making process for addressing the disruptive effect of inclement weather on the traffic system. These tools were developed, applied, calibrated and tested in three different locations across the US. The networks selected include Salt Lake City, UT, New Yorks Long Island Expressway Area and Chicago, IL. In addition to building the networks and developing the various demand and supply inputs, calibration of the traffic models to local weather conditions has provided a rich database of calibrated weather-sensitive traffic models that could be applied in any region. The methodology is configured to run with real-time traffic data (e.g. from loop detectors or RTMS sensors). The methodology was used to evaluate the effectiveness of different weather-related traffic management strategies in specific areas. The project has successfully demonstrated the potential of a TrEPS installation that is always on, up to date and adaptively calibrated to address a range of weather conditions and scenarios, retrievable on demand by local operating agencies through a scenario manager.