Line control systems on urban freeways use variable message signs for traffic dependent speed limits, no passing zones, lane closing and warning such as wet or icy road. Due to the high access density in urban environment the line control systems must be completed by ramp metering facilities. The metering rate is controlled by occupancy data from the main stream and from the on ramp traffic as well as from the congestion situation in the access area. The large variety of displays and metering rates leads to a system of up to 60 different thresholds. Embedded in a decision logic with deterministic as well as with fuzzy control elements, the thresholds are chosen by an expert system based on the knowledge of the effects of the provisions for traffic flow control. The knowledge-based optimisation for threshold choice uses a traffic flow model for calculating an objective function with optimum safety and throughput strategy under given boundary conditions. It also profits from a knowledge acquisition expertise, which props on the experience of consultant engineering in designing and operating line control systems. The optimisation procedure is shown for 2 practical applications. One is a line control system on the Autobahn A3 near Nurnberg, Germany. This system is in an automatic operation mode and yields data for completely assessing the decision logic and the optimisation procedure. The second example is a design study for the CITRAC / FEDICS traffic control project in Glasgow, Scotland, which shows the applicability of the derived techniques for the design of advanced traffic control systems. (A)
Samenvatting