DocumentCode :
515158
Title :
Traffic flow prediction based on wavelet transform and Radial Basis Function network
Author :
Yang, Wen ; Yang, Dongyuan ; Zhao, Yali ; Gong, Jinli
Author_Institution :
Sch. of Transp. Eng., Tongji Univ., Shanghai, China
Volume :
2
fYear :
2010
fDate :
9-10 Jan. 2010
Firstpage :
969
Lastpage :
972
Abstract :
Exact prediction of traffic flow is the key technology of traffic flow guidance and traffic system management. A kind of wavelet neural network model combined with the advantages of wavelet transform and RBF network was presented for short-term traffic flow prediction. After the wavelet decomposition and reconstruction were made to traffic flow data with similar periods, signal components respectively were predicted by RBF neural network, and prediction results were synthesized. Furthermore, different time intervals were adopted for prediction and effects were compared with each other according to several evaluation indexes. The results show that prediction effect is better than just predicting by neural network in prediction precision and network convergence. Therefore, there are favorable prospects for applications.
Keywords :
radial basis function networks; traffic engineering computing; wavelet transforms; RBF network; radial basis function network; traffic flow guidance; traffic flow prediction; traffic system management; wavelet neural network; wavelet transform; Convergence; Network synthesis; Neural networks; Predictive models; Radial basis function networks; Signal synthesis; Technology management; Telecommunication traffic; Traffic control; Wavelet transforms; Flow Prediction; RBF Neural Network; Traffic Flow; Wavelet Transform;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Logistics Systems and Intelligent Management, 2010 International Conference on
Conference_Location :
Harbin
Print_ISBN :
978-1-4244-7331-1
Type :
conf
DOI :
10.1109/ICLSIM.2010.5461098
Filename :
5461098
Link To Document :
بازگشت