Chen Ning
School of Mechanical and Automotive Engineering, Zhejiang University of Science and Technology, Hangzhou, China
Wen Xiaoyue
Zhejiang Enjoyor Electronics Co. Ltd., Hangzhou, China
Qu Xiao
School of Automation and Electrical Engineering, Zhejiang University of Science and Technology, Hangzhou, China
ABSTRACT
It is inevitable that the defectiveness of urban traffic flow data always occur in collecting information due to the sensors failure. In order to mend those defective data, a new fitting method based on SARBF neural networks for defective-data-mending of urban traffic flow is presented in this study. It is not only an approach to analyzing the traffic data based on spatial autocorrelation but also a method on mending the defective data based on the RBF neural networks fitting technique. Firstly, the effectiveness of the defective-data-mending for complete data is great improved by using the spatial autocorrelation according to the urban traffic grid. Secondly, not only the mending precision but also the limitation of regression analysis is developed because of using RBF neural network. The experiment was held to in Hangzhou city. It is shown by the experiments results that the fitting method brought up in this study is quite practicable to mend the defective data of urban traffic flow.
PDF References Citation
How to cite this article
Chen Ning, Wen Xiaoyue and Qu Xiao, 2013. Reasearch on Fitting Method for Defective-data-mending of Urban Traffic Flow Based on SARBF Neural Networks. Information Technology Journal, 12: 8570-8575.
DOI: 10.3923/itj.2013.8570.8575
URL: https://scialert.net/abstract/?doi=itj.2013.8570.8575
DOI: 10.3923/itj.2013.8570.8575
URL: https://scialert.net/abstract/?doi=itj.2013.8570.8575
REFERENCES
- Zhang, H., W. Wang and H.Z. Gu, 2005. Application of cluster analysis and stepwise regression in predicting the traffic volume of lanes. J. Southeast Univ. (English Edn.), 28: 359-362.
Direct Link