Paper Details

TRB 91st Annual Meeting (January 22-26, 2012)

Event Number:	648
Event Title:	Highway Safety Performance
Event Date:	Jan 24 2012 7:30PM- 9:30PM
Event Location:	Marriott, Salon 2
Event Description:
Event Agenda:

Crash Frequency Analysis for Urban Expressways by Considering Segment Type (12-1965)

Incidents occurring on congested urban expressways can have enormous impacts in terms of capacity reduction and commuter time loss. Minor crashes are usually neglected in the field of traffic safety, but should be considered from the perspective of congestion alleviation. This paper investigated crash frequency extracted from incident data obtained by video monitoring system, which is implemented on three expressways in Shanghai. Features of urban expressways are quite different from that of freeways, especially in length of segments between two consecutive ramps. In general, only ramp density was considered when analyzing the effect of ramps on crash occurrence and all segments were modeled together, but the differences among different types of segments in terms of varied ramp junction areas were ignored. In this analysis, disaggregate data was used by considering variation by day of week and peak/off-peak hours. Biased structure in annual aggregate data was avoided and the effect of temporal aggregation was then discussed. Segments are classified into four categories, namely merge segment, diverge segment, weaving segment and basic segment, and prediction models were established for each type of segment respectively. Results showed that temporal aggregation of data led to loss of impact factors and change in effect size. It was also found that relationship between crash frequency and its impact factors varied remarkably among different segment types. The relationship between crash frequency and segment length is not constantly linear and segment length has a significant two-side effect on crash occurrence on expressways.

Authors
     Tu, Yingfei , Tongji University, China
     Zhang, Jiangman , Tongji University, China
     Yang, Chao , Tongji University, China
     Chen, Xiaohong , Tongji University, China