Optimization Method of Inspection Point Capacity on
V. M. Antonova 1, D. O. Volkov
2, N. A. Grechishkina 3, N. A. Kuznetsov 3
Bauman Moscow State Technical University,
Baumanskaya 2-ya, 5, Moscow 105005, Russia
Moscow Institute of Physics and Technology (State University),
MIPT, 141701, Moscow region, Dolgoprudnyy, Institutskii lane, 9
3 Kotel'nikov institute of radio engineering and electronics of RAS,
Mokhovaya 11-7, Moscow,
The paper is received on March 16, 2017
Abstract. Ever-increasing demands for passenger
safety on high-speed transport (e.g. underground railway, aeroexpress trains,
commuter trains, etc.) have resulted in installing screening systems at the
passenger checkpoints, the systems having to work with the necessary throughput
considering passenger traffic unevenness and not gathering long queues. This
leads to a number of challenges related to managing users’ access to inspection
point boxes. The aim of the given work is to build a model for predicting
values of an objective function on the basis of several input variable values
using the classification method of incoming passenger traffic based on a
decision tree. This method is demonstrable and doesn’t require data
preprocessing for studying. The main feature of passenger traffic being its
unevenness, i.e. its time-to-time variability (according to time, days of week,
seasons), the spread of factors influencing passenger traffic unevenness is
rather large; this results in a great number of their various combinations.
Thus more than ten thousand different incoming distributions of the passenger
traffic according to the time of the day has been received as a result of the
statistic simulation which presents a double-humped distribution carried out by
means of combining three normal distributions having peaks in the morning,
afternoon and evening. Applying a decision tree to the first hour of inspection
point operation allows to define the capacity of incoming passenger traffic
during the whole day and therefore the necessary quantity of inspection points
can be determined. This leads to reducing queues to the inspection points. Similar
problems can be solved by mobile operators who serve subscribers arriving at
the railway stations and airports.
decision tree, passenger
traffic, point boxes.
Stepanov S. N.
Theory of teletraffic: concepts, models, applications. Moscow: Hot line –
telecom Publ., 2015 – 868 p.
Antonova V. M.; Kuznetsov N. A.; Starostenko A. M.; Volkov D. O. Solving the
Classification Problem for Building Predicative Models of Passenger Traffic in
the MATLAB Environment.
Information processes. 2017. (In
3. Antonova V.
M.; Bogomolov I.V.; Selenina A. L. The technique for analyzing large amounts of
data using the example of a classification problem in the design of heating
systems. The development of modern science: theoretical and applied aspects of a
collection of articles by students, undergraduates, graduate students,
scientists and teachers (03.29.2016). Perm, 2016. P. 17-21. (In Russian)
Advantages in the LTE picocells networks. Proceedings of the North Caucasus
Branch of the Moscow Technical University of Communications and Informatics.
2016. Т. 1.
p. 27-29. (In Russian)
V.M.Antonova, D.O.Volkov, N.A.Grechishkina, N.A.Kuznetsov.
Optimization method of inspection point capacity on
Zhurnal Radioelektroniki - Journal of Radio Electronics,
2017, No. 3. Available at http://jre.cplire.ru/jre/mar17/7/text.pdf.