Highlights Difference between the expected and observed travel times is ≤15% for 90% of the samples. ARIMA model has the potential to forecast relative changes in travel time, irrespective of the incident condition. Modeling using expected travel time yields marginally better results than minimum travel time. Both, upstream and downstream segments influence the target segment travel time.
Abstract This paper focuses on forecasting short-term spatiotemporal relative change in travel time (RCTT) on a freeway corridor. The RCTT was used instead of the travel time to capture the relative effect (say) due to a vehicle accident when compared with the no incident condition. Two types of RCTT were considered: travel time/expected travel time and travel time/minimum travel time. Databases were developed using data for 135 “Vehicle Accident” affected days and 128 days when there were no incidents, on a freeway corridor. The expected travel time was computed through averaging of all the travel time samples for a specific time period. The percent difference between the expected travel time and the observed travel time is less than 15% for 90% of the samples, indicating that the computed expected travel time represents the travel time during no incident condition. Autoregressive Integrated Moving Average (ARIMA) model was then applied to model RCTT. The Mean Absolute Percent Error (MAPE) of the fitted models, as well as for the validated models, was less than 15% for most of the segments, indicating the effectiveness of ARIMA in forecasting short-term spatiotemporal RCTT due to a vehicle accident.
Forecasting short-term relative changes in travel time on a freeway
2019-03-14
13 pages
Aufsatz (Zeitschrift)
Elektronische Ressource
Englisch
Microsimulation-Based Framework for Freeway Travel Time Forecasting
Transportation Research Record | 2019
|Reliable Real-Time Framework for Short-Term Freeway Travel Time Prediction
British Library Online Contents | 2006
|Reliable Real-Time Framework for Short-Term Freeway Travel Time Prediction
Online Contents | 2006
|Freeway Short-Term Travel Time Prediction Based on Dynamic Tensor Completion
Transportation Research Record | 2019
|A space–time diurnal method for short-term freeway travel time prediction
Online Contents | 2014
|