Ramp metering (or on-ramp traffic restriction) aims to optimize traffic conditions in the vicinity of a motorway merging section. However, to prevent excessive on-ramp queues from disrupting adjacent local road network, they have to be managed properly. Traditionally, occupancy measurements from on-ramp induction loops were used to activate queue-over-ride operation when certain threshold is exceeded. One of the aims of the European project RHYTHM was to assess the benefit of utilizing video detection technology for on-ramp queue management of ramp metering. A new queue control algorithm integrated into ALINEA strategy was developed to utilize the on-ramp queue length information provided by video detection. A microscopic simulation model (FLOWSIM) was used for evaluating the new ramp metering prototype. The test site was a model of the Mnchen-Messe Riem junction at the A94 near Munich, Germany which is the RHYTHM real test site. Three levels of ramp demands were used (500, 600 and 700veh/h) and several simulation runs with different random seeds were undertaken to ensure the reliability of results. The evaluation used a number of assessment measures including total vehicle travel time and the time of on-ramp queue valuation to maximum ramp length. The traffic condition without metering was considered as a reference base and the new prototype was compared with two other metering situations: (i) ALINEA without any on-ramp queue control. (ii) ALINEA with simple queue control using induction loops. the impact of ramp metering was reduced significantly when ramp queues were considered. However, without on-ramp queue control, the time of queue valuation to maximum storage space was very large (more than 50% of the time) so that the benefit of ramp metering in such case is likely to be void. The larger the ramp demands the lower the impact of ramp metering with queue control as on-ramp queues build up quickly activating queue-over-ride more often. The total vehicle travel time was reduced by 24% (positive impact) when ramp demand was 500veh/h while it increased of 2% (negative impact) when ramp demand was 700veh/h. The assessment has also shown that video detection and the new on-ramp queue management algorithm has contributed to better results compared to traditional loop-occupancy approach. It helped to reduce the time of queue valuation to maximum storage space and to improve the impact of ramp metering when ramp demand was appropriate. For the covering abstract please see ITRD E135207.
Samenvatting