PROBABILISTIC NATURE OF BREAKDOWN PHENOMENON AND ON-RAMP METERING IN THREE-PHASE TRAFFIC THEORY

Problems and limitations of freeway traffic control associated with probabilistic nature of freeway capacity at a freeway bottleneck are considered. Specifically, the probabilistic feature of traffic breakdown at an on-ramp bottleneck leads to a great limitation for reliable applications of free flow control approach in which free flow should be maintained at the bottleneck. For this reason, the author has recently introduced a congested pattern control approach (ANCONA), in which congestion at the bottleneck is allowed to set in. Only after traffic breakdown occurs randomly at a bottleneck, ANCONA begins to perform. ANCONA is able to keep congestion conditions at the minimum possible level at the bottleneck. In particular, the congested pattern does not propagate upstream. In the paper, ANCONA approach is applied for traffic control of congestion occurring at two adjacent bottlenecks, off- and on-ramp bottlenecks. It is shown that if without on-ramp metering a congested pattern occurring at the downstream off-ramp bottleneck propagates subsequently upstream, due to ANCONA application upstream propagation of congestion can be stopped. As a result, congestion is spatially localized. In addition, congestion is kept at the minimum possible level at the bottleneck.