PT Journal
AU Mohanty, M
   Ranjan Samal, S
   Samal, K
TI Developing a Speed-Based Congestion Severity Index Using the Clustering Technique for Developing Countries
SO Communications - Scientific Letters of the University of Zilina
PY 2025
BP D67
EP D75
VL 27
IS 2
DI 10.26552/com.C.2025.020
WP https://komunikacie.uniza.sk/artkey/csl-202502-0005.php
DE traffic congestion; percentile speed; k-mean clustering; silhouette value; congestion severity index
SN 13354205
AB A novel approach for traffic congestion assessment has been presented using percentile speeds as key indicators, focusing on urban roads. By evaluating the 98th, 85th, and 15th percentile speeds, authors of the research developed a congestion severity index, offering a more precise and intuitive method for analyzing traffic flow compared to traditional travel time-based indices. Key congestion indices, such as the Planning Time Index (PTI) and Travel Time Index (TTI) were compared to percentile speeds, revealing a significant association with the 15th and 85th percentile speeds. The K-means clustering technique was applied to classify congestion severity into three levels, validated by a high silhouette value indicating the robust clustering. The study's speed-based congestion severity index provides a practical and efficient framework for real-time congestion management, particularly in heterogeneous traffic environments.
ER