Publication Details

Estimation of missing values in traffic density maps

PETRLÍK, J.; KORČEK, P.; BESZÉDEŠ, M.; FUČÍK, O.; SEKANINA, L. Estimation of missing values in traffic density maps. 8th Doctoral Workshop on Mathematical and Engineering Methods in Computer Science. Brno: 2012. p. 1 (1 s.).
Czech title
Odhad chybějících hodnot v dopravních zátěžových mapách
Type
abstract
Language
English
Authors
Petrlík Jiří, Ing., Ph.D. (RG EHW)
Korček Pavol, Ing., Ph.D. (DCSY)
Beszédeš Marián
Fučík Otto, doc. Dr. Ing. (DCSY)
Sekanina Lukáš, prof. Ing., Ph.D. (DCSY)
Keywords 
Optimization and Control: Theory and Modeling,Statistical Modeling, Data Mining and Analysis
Abstract

This contribution is based on the paper: Petrlík, J., Korček, P., Fučík, O., Beszédeš, M., Sekanina, L.: Estimation of missing values in traffic density maps , In: Proc. of the 15th International IEEE Conference on Intelligent Transportation Systems, 2012, accepted. 

Annotation
The traffic density map (TDM) represents the density of road network traffic as the number 
of vehicles per a specific time interval. This interval can be given in minutes 
or hours. Usually, TDMs are used by traffic experts 
as a base documentation for planning a new infrastructure (long-term) or 
by drivers for showing a current traffic status (short-term). Such TDMs can be composed
automatically -- with the aid from standard surveillance technologies 
(e.g. various data sensors such as loop detectors or traffic cameras).
Another approach, which can be used for TDM calcultion, is the manual counting on selected 
road segments. In the situation 
where it is not possible to cover the whole traffic network with the field data, 
missing areas must be completely excluded from the traffic density estimation. To avoid this situation we propose two methods for 
estimation of missing density values.
TDM can be viewed as a directed graph, where each node n represents a crossroad 
and each edge represents a particular road segment. The density on the edge, d_e, 
represents the number of incoming our outgoing vehicles per time interval 
on a given edge e. Typically 40 % of
values are missing and have to be estimated. The historic value of density h_e (e.g. measured some time ago)
 can also be available for some undefined edges.
In the first method for estimation of missing values, the problem is formulated relatively strictly in terms
of quadratic programming (QP) and a QP solver is utilized to find a solution. The second, more general method is based on
a multiobjective genetic algorithm which allows us to find a reasonable compromise among several objectives that a traffic expert
may formulate. These two methods can work automatically or they  can be used by a traffic expert for an iterative density estimation.
The methods provide similar quality of results when evaluated and compared using synthetic and real-world data (center of Prague).
Published
2012
Pages
1–1
Book
8th Doctoral Workshop on Mathematical and Engineering Methods in Computer Science
Place
Brno
BibTeX 
@misc{BUT192852,
  author="Jiří {Petrlík} and Pavol {Korček} and Marián {Beszédeš} and Otto {Fučík} and Lukáš {Sekanina}",
  title="Estimation of missing values in traffic density maps",
  booktitle="8th Doctoral Workshop on Mathematical and Engineering Methods in Computer Science",
  year="2012",
  pages="1--1",
  address="Brno",
  note="abstract"
}
Back to top