Multistage Fuzzy Logic Controller for Expressway Traffic Control during Incidents

Abstract

In this research study, a multi-stage Fuzzy Logic Controller (MS-FLC) is developed for traffic control for incident management on expressways. The MS-FLC serves as the traffic operator's decision-making support tool at the operational level.The MS-FLC gathers real-time traffic and incident data in order to analyze and predict traffic conditions as well as to suggest alternative control measures to the traffic operator in the form of linguistic expressions. The MS-FLC is embedded in a traffic simulator controller (TSC) prototype and is evaluated by comparing its performance with no control scenario and ALINEA\Q, a popular local ramp control algorithm, across several incident scenarios in a simulation environment. In general, the MS-FLC outperforms ALINEA\Q with respect to global objectives. In particular, whereas the ALINEA\Q algorithm favors the mainline, the MS-FLC algorithm significantly improves mainline travel conditions while substantially reduces ramp queues. It is concluded that, if properly designed the MS-FLC serves as a robust tool for traffic control on expressways under incident conditions.

Developing a fuzzy-based decision-making procedure for traffic control in expressway congestion management

“…Herein, we are motivated to conduct a broader research study that develops a multi-stage fuzzy logic controller (MS-FLC) for traffic control in congestion management on expressways. The MS-FLC development roadmap starts from the local control to corridor-wide control, for which the performance of initial development has been successfully evaluated and the results presented in [12]. The MS-FLC consists of 3 stages: (1) evaluation of existing traffic congestion; (2) prediction of incoming traffic congestion; and (3) recommendation of control strategies and control actions.…”

“…The rule conditions are joined with AND connectives. Under some specific favourable mainline conditions (rules 1, 2, 7, 8, 13, 14) and critical mainline congestion (rules 6,12,18,23,24) the rule are triggered without considering the ramp queue. In correspondence to the control objectives, the conditions of the rules consider the traffic condition (congestion level, CL) upstream of the incident (downstream of the ramp), the traffic demand (indicated by the * C V ratio) upstream of the ramp, and the ramp queue (see Fig.…”

“…In addition, the ramp flow is adjusted according to the ramp queue status so as to maintain acceptable ramp queue (rule 4), to prevent excessive ramp queue (rules 5,11,21,22), or to maintain a balance between objectives (rules 8, 13,14,19). Finally, if the traffic on the mainline is congested, the restriction of the ramp flow is to target preventing a secondary ramp queue at the ramp bottleneck (rules 12,15,16,17). The reason for this restriction is that when the mainline is congested, the ramp traffic will hardly find an acceptable gap to join the mainline, so a secondary queue of the metered vehicles may form spontaneously.…”

“…As stated in the "Introduction", in the broader research project [12], we have developed a multi-stage fuzzy logic controller (MS-FLC) for traffic control for congestion management on expressways. The decision-making process for traffic control for congestion management on expressways using the MS-FLC consists of three tasks: (1) evaluation of current traffic congestion; (2) prediction of traffic congestion tendency; and (3) recommendation of control strategies and control actions to alleviate congestion.…”

“…Common control measures that are employed in expressway networks include ramp control, link control, and driver information and guidance systems [3,10]. Ramp control applies devices such as metering signals to regulate the number of vehicles entering the expressway [11,12]. Ramp control is the most efficient and popular measure to control expressway traffic [3,7,10,13,14].…”