A Functional Approach to Monitor and Recognize Patterns of Daily Traffic Profiles
Abstract
Functional Data Analysis (FDA) is a collection of statistical techniques for the analysis of information on curves or functions. This paper presents a new methodology for analyzing the daily traffic flow profiles based on the employment of FDA. A daily traffic profile corresponds to a single datum rather than a large set of traffic counts. This insight provides ideal information for strategic decision-making regarding road expansion, control, and other long-term decisions. Using Functional Principal Component Analysis the data are projected into a low dimensional space: the space of the first functional principal components. Each curve is represented by their vector of scores on this basis. The principal component scores are used for clustering and also to identify outliers (meaning that there was a bad performance in the recording of data or special circumstances affected the traffic) and to monitor the traffic profile by multivariate control charts. This paper introduces this new methodology and illustrates good results by using 1-min traffic data from the I-94 Freeway in the Twin Cities, Minnesota (U.S.) metroplex ranging from 2004 to 2011.
Recommended Citation
I. G. Guardiola et al., "A Functional Approach to Monitor and Recognize Patterns of Daily Traffic Profiles," Transportation Research Part B: Methodological, vol. 65, pp. 119 - 136, Elsevier, Jul 2014.
The definitive version is available at https://doi.org/10.1016/j.trb.2014.04.006
Department(s)
Engineering Management and Systems Engineering
Keywords and Phrases
Clustering; Functional data analysis; Multivariate control charts; Temporal traffic flow
International Standard Serial Number (ISSN)
0191-2615
Document Type
Article - Journal
Document Version
Citation
File Type
text
Language(s)
English
Rights
© 2014 Elsevier, All rights reserved.
Publication Date
01 Jul 2014