Ain Shams Engineering Journal (Jun 2024)
Traffic congestion evaluation of urban streets based on fuzzy inference system and GIS application
Abstract
This paper satisfies the requirements for reliable and inexpensive congestion detection in urban road networks. The objective of this research is to use fuzzy logic to detect the traffic conditions states based on sets of rules that compare the filed traffic states. An alternative approach that enables knowledge based on effective and efficient methods of detecting traffic congestion. The main strategies in this work include the detection of congestion based on index measures of traffic speed, index speed reduction, and speed ratio. Spatial analysis of traffic data utilizing ArcGIS application to produce digitized street maps of congestion assessments through GIS traffic data. Utilizing the Fuzzy Inference System (FIS) approach for adopted traffic parameters provides an analytical solution for ambiguous and uncertain problems. The categories of traffic input parameters distinguish states of congestion levels through the determination of values of congestion index levels. The analysis of the speed reduction index illustrated the hot spots within the study network which represent Bab Al-Moathum zone; Al-Mawal (Mustansiriyah University zone); Al-Kindi government hospital induced heavy congestion with an index range of (0.651–0.717). A precise threshold to describe the level of congestion using the Fuzzy Inference System to evaluate urban streets into different categories; Free-flow, Normal, Moderate, Heavy congested, and Blocked. A contribution of two input traffic parameters is considered to quantify congestion in one output that combines different congestion field measures. Over 15 links (Bab Al-Moathum zone, Palestine Street near Mustansiriyah University, Al-Kindi govern ate hospital with black color induced worse traffic conditions, resulting in a blocked effect. The rest of the segments within the case study range from heavily congested to normal. A more realistic and detailed view of traffic congestion for selected street networks is obtained based on a fuzzy inference system as to traditional methods that consider one parameter for traffic performance.