Izzah Syahirah Binti Mohamad Arif Universiti Pertahanan Nasional Malaysia
Understanding the spatial and temporal behaviour of road traffic volume
is essential for effective transportation planning and traffic monitoring,
particularly in rapidly urbanising regions such as Klang, Malaysia. This
study employs network theory to assess the importance of road traffic
volume census stations based on their temporal traffic patterns. This study
focuses on eleven Road Traffic Volume Malaysia (RTVM) census stations
in the Klang area, Selangor, using hourly traffic volume data collected over
a seven-day period.
The analysis is conducted in two main parts. Part 1 of the Thesis
investigates the role of RTVM Census Stations through a network analysis
methodology. Three types of Networks have been developed: an undirected
correlation based network, which shows the similarities between the
Traffic Volume Patterns for each station; a directed unweighted network,
which describes directional relationships between Stations; and a directed
weighted network, which reflects the volume of traffic between stations.
Centrality Measures such as Degree Centrality, Betweenness Centrality and
Closeness Centrality are then employed to help identify key stations within
a particular network structure.A combined station importance scoring
framework is then employed to synthesise results obtained from the
different network representations.
In Part II, short-term traffic volume forecasting is performed using basic
statistical time series models. According to the findings of the present
analysis, SARIMA has an advantage, in that, on average, it provides a more
accurate prediction of traffic patterns than ARIMA. The overall results
from the network analyses show that stations BR101 have consistently
emerged as the two most important census stations, regardless of which
network configuration has been tested. This study addresses limitations in
existing traffic monitoring practices by providing a systematic approach
to evaluating station importance and traffic behaviour. The findings offer
practical insights for traffic monitoring optimisation and support data-driven
decision-making in transportation planning within Klang, Malaysia.
