Signal Violation Effects on Pedestrian Delay at Signalized Intersections

Pedestrian delay must be considered when evaluating pedestrian level of service at signalized intersections. Pedestrian delay is usually modeled with the assumption that pedestrian arrivals are random and pedestrians enter the crosswalk only during green interval. However, it is often observed that pedestrians accept gaps and cross during red time to minimize their own delay. A field study has been carried out in residential areas in the downtown city of Shanghai, China. A total of 92 cycles with 1677 pedestrians at 3 crosswalks have been observed. Delays experienced by pedestrians were found to be 1% to 80% lower than those that would be predicted with complete signal compliance under different traffic situations. Pedestrian arrival and departure characteristics have been precisely analyzed, and a new delay model based on "arrival-departure curves" has been developed and calibrated. Besides pedestrian compliance rate of those who arrive during non-green interval, other parameters representing pedestrian gap acceptance behavior and proceeding features during red time have been introduced, which are influenced by saturation degree of lanes at entrances and exits, availability of gaps between consecutive phases, or between queuing vehicles and platoons from upstream intersections etc. The model has been validated using the field data and a comparison of measured delay and estimated delay provided by different models has been carried out. The results indicated that the new model could estimate pedestrian delay, and the method was widely applicable.