DocumentCode :
2942516
Title :
The High Frequency Traffic Flow Analysis
Author :
Zhang Gaoyu ; Zhou Zhizhao ; Zhou Huan
Author_Institution :
Comput. Sci. Inst., Fudan Univ., Shanghai, China
Volume :
2
fYear :
2009
fDate :
12-14 Dec. 2009
Firstpage :
221
Lastpage :
224
Abstract :
The statistical analysis of traffic flow is the fundamental of road traffic forecasting with data mining. In the common traffic flow theory, macroscopic flow model and data presentation shows chaos and stochastic distribution. However, in the microscopic based traffic model, it is revealed that in high frequency sampled test, the variables such as flow, density, and lane usage have common trend of U shape in daytime, e.g. high in peak time and low in off-peak time. Based on the traffic database of cross-river channel in LuJiaZui, PuDong, Shanghai, the statistics method is used for different sampling frequency to validate the mean, standard variance, and kurtosis. The variant ratio is used to analyze the first order negative correlation among high frequency sampled traffic data. The result shows the models (e.g. ARMA model) used in low frequency sampling data cannot directly apply to high frequency samples. In order to validate the U shape distribution of traffic data in daytime, the wavelet based nerve network is used for the analysis of `calendar phenomena´ of cross-river tunnel traffic.
Keywords :
data mining; road traffic; statistical analysis; stochastic processes; traffic engineering computing; ARMA model; U shape distribution; calendar phenomena; cross-river tunnel traffic; data mining; data presentation; high frequency traffic flow analysis; macroscopic flow model; microscopic based traffic; road traffic forecasting; statistical analysis; stochastic distribution; traffic database; wavelet based nerve network; Chaos; Data mining; Frequency; Roads; Sampling methods; Shape; Statistical analysis; Stochastic processes; Telecommunication traffic; Traffic control; high frequency; nerve network; traffic flow; wavelet;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Computational Intelligence and Design, 2009. ISCID '09. Second International Symposium on
Conference_Location :
Changsha
Print_ISBN :
978-0-7695-3865-5
Type :
conf
DOI :
10.1109/ISCID.2009.202
Filename :
5371078
Link To Document :
بازگشت