Title :
Short-term travel time estimation: A case study
Author :
Mohammed, Emad A. ; Aulakh, Charan ; Krishnamurthy, Diwakar ; Naugler, Christopher T. ; Far, Behrouz H.
Abstract :
This paper presents a case study of traffic travel time estimation for the Deerfoot Trail highway in Calgary, AB, Canada. We analyzed the travel time for 12 hours per day for 5 consecutive months. The collected data are posted by Google maps, which are collected using hypertext transfer protocol requests. This paper provides a construction of the Holt-Winters forecasting model that streamlines the estimation of travel time for a specific hour of the day. The Holt-Winters model with exponential smoothing average has the best estimate of travel time. The model goodness of fit for the data collected at 07:00AM shows that the multiplicative Holt-Winters model has the best performance (mean square error 15.733 with R2 value of 0.8435). A fleet driver can use the results to decide to travel at a given hour. Moreover, decision makers can use the results to plan for road events.
Keywords :
intelligent transportation systems; mean square error methods; transport protocols; Deerfoot trail highway; Google maps; Holt-Winters forecasting model; ITS; MSE; decision makers; exponential smoothing average; fleet driver; hypertext transfer protocol requests; intelligent transportation system; mean square error; multiplicative Holt-Winters model; road events; short-term travel time estimation; Additives; Data models; Estimation; Google; Mathematical model; Predictive models; Time series analysis; ARIMA; Holt-Winters; Time Series Analysis; Traffic Pattern Analysis; Travel Time Estimation;
Conference_Titel :
Electrical and Computer Engineering (CCECE), 2015 IEEE 28th Canadian Conference on
Conference_Location :
Halifax, NS
Print_ISBN :
978-1-4799-5827-6
DOI :
10.1109/CCECE.2015.7129324
