In a typical traffic time series, a within-day seasonal cycle is evident from the similarity of traffic data from one day to the next; a within-week seasonal cycle can be observed when comparing traffic data on corresponding days of adjacent weeks. Recently a new method called the double seasonal (DS) Holt-Winters method has been proposed to deal with double seasonality in time series. The primary objective of this study is to investigate the possibility of using the DS method to account for both within-day and within-week seasonal cycles in traffic time series. Based on speed data from Freeway I-5, California, the DS method, the standard Holt-Winters method and the ARIMA method are employed to perform short-term traffic forecasting. The result shows that the proposed DS model outperforms the other two methods, indicating that the double seasonal Holt-Winters approach is a promising and effective method for short-term traffic forecasting.
Short-Term Traffic Forecasting Using the Double Seasonal Holt-Winters Method
16th COTA International Conference of Transportation Professionals ; 2016 ; Shanghai, China
CICTP 2016 ; 162-173
2016-07-01
Conference paper
Electronic Resource
English
Short-term Forecast Model of Vehicles Volume Based on ARIMA Seasonal Model and Holt-Winters
DOAJ | 2017
|Prediction Scheme of Railway Passenger Flow Based on Multiplicative Holt-Winters Model
Trans Tech Publications | 2013
|Adaptive Seasonal Time Series Models for Forecasting Short-Term Traffic Flow
Transportation Research Record | 2007
|Adaptive Seasonal Time Series Models for Forecasting Short-Term Traffic Flow
Online Contents | 2007
|Short-term Traffic Forecasting using Neuronal Network
British Library Conference Proceedings | 1995
|