A knowledge-based decision support system for incident traffic congestion management

Abstract

A Knowledge-based Decision Support System (KB-DSS) based on a multi-stage Fuzzy Logic Controller (MS-FLC) is developed for traffic congestion management on expressways. The MS-FLC receives real-time traffic and incident data to analyse and anticipate the traffic conditions, to recommend alternative control measures in the form of natural languages for the human operator to select control decisions, and to calculate control settings to manage traffic congestion. In a case study, the KB-DSS is evaluated on a simulated network in comparison to ALINEA\Q, a popular ramp control method, across various traffic and incident situations. The results showed that: (i) the KB-DSS provides a systematic procedure in deriving control actions and a good capability to deliver linguistic expressions; (ii) the KB-DSS outperforms ALINEA\Q with respect to global objectives across many scenarios, attains significant improvements of mainline travel conditions and substantial reductions of ramp queues. These advantages make the KB-DSS a robust tool for traffic control for incident congestion management on expressways.