Abstract
Short-term prediction is very important for real-time traveler information systems. While many current applications provide mean travel times, not many provide reliability metrics such as travel time reliability. In addition, the mean corridor or route travel time is usually estimated by simply summing link travel time means, which does not account for the correlation between link travel times. To address these two issues, this paper studies corridor travel time reliability prediction at a 15-minute level under various traffic conditions using real-time Bluetooth travel time data. This data is reduced to both link-based and corridor-based travel time formats. Prediction models for corridor travel time reliability, using link travel times and corridor travel times, are developed and compared. The nonlinear autoregressive with exogenous inputs (NARX) Model was found to be the best approach because the predicted reliability intervals of arrival time were the most accurate. The link-based NARX model was found to yield comparable results to the corridor-based NARX model.
| Original language | English (US) |
|---|---|
| State | Published - 2014 |
| Event | 21st World Congress on Intelligent Transport Systems: Reinventing Transportation in Our Connected World, ITSWC 2014 - Detroit, United States Duration: Sep 7 2014 → Sep 11 2014 |
Other
| Other | 21st World Congress on Intelligent Transport Systems: Reinventing Transportation in Our Connected World, ITSWC 2014 |
|---|---|
| Country/Territory | United States |
| City | Detroit |
| Period | 9/7/14 → 9/11/14 |
Keywords
- Bluetooth data
- Nonlinear AutoRegressive with eXogenous inputs model (NARX)
- Real-time prediction
- Travel time reliability
ASJC Scopus subject areas
- Control and Systems Engineering
- Mechanical Engineering
- Automotive Engineering
- Transportation
- Electrical and Electronic Engineering