Development of Congestion Severity Index for Uncontrolled Median Openings Utilising Fundamental Traffic Parameters and Clustering Technique: A Case Study in India

• Published in: International journal of ITS research Vol. 21; no. 3; pp. 461 - 472
• Main Authors: Mohanty, Malaya, Sarkar, Bitati, Pattanaik, Madhu Lisha, Samal, Satya Ranjan, Gorzelańczyk, Piotr
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.12.2023; Springer Nature B.V
• Subjects: Automotive Engineering; Civil Engineering; Clustering; Computer Imaging; Crashes; Electrical Engineering; Engineering; Free flow; Pattern Recognition and Graphics; Roads; Roads & highways; Robotics and Automation; Time measurement; Traffic congestion; Traffic flow; Traffic volume; Travel time; User Interfaces and Human Computer Interaction; Vision; Travel time index; Clustering; Median opening; Planning time index; Traffic congestion
• ISSN: 1348-8503; 1868-8659
• DOI: 10.1007/s13177-023-00365-1

Traffic congestion has harmful implications on road users and leads to rise in both commuting time and the tendency of road rage, increasing the frequency of road crashes. At median openings (MO), due to complex movement of U-turns and approaching traffic, measurement and assessment of congestion is imperative. Identifying the level of congestion at median openings can help in strategizing location specific steps to ease the problem. In this study, the travel time reliability measures like planning time index (PTI) and travel time index (TTI) have been modified with relation to median opening by incorporating steady speed as compared to free flow speed used earlier. Video graphic analysis was conducted and speed data is extracted from 4500 vehicles travelling at different traffic volumes. PTI and TTI have been correlated with 95th, 85th, and 15th percentile speeds and have been used to develop the ranges of congestion severity index. It is observed that the congestion indices alone cannot justify the amount of congestion since their values saturated after a threshold traffic volume of 3000–4000 vph. 15th percentile speed along with the 85th and 95th percentile speeds gives a better picture of congestion since they encompass the slow moving vehicles too. The percentile speeds have been utilized to designate the congestion severity index by using k-mean clustering. Traffic situations where 95th, 85th, and 15th percentile speeds are below 39.6, 35.6, and 22.7 kmph, it can be attributed to severe traffic congestion whereas the same percentile speeds if are observed to be above 52, 46, and 31 kmph, the traffic flow can be represented as free flow/no sign of any congestion. The present study helps in categorizing traffic congestion at median openings in a rational way, thereby undertaking proper steps and actions to alleviate them. However, the results obtained are valid for median openings on 6-lane divided urban roads and the same methodology can be conducted for roads of different lane configuration for better understanding.