Evaluating Location of Placement and Spacing of Detectors for Real-Time Crash Prediction on Urban Expressways

摘要: The concept of real-time crash prediction is in its early stage and so far the focus has been on evaluating different methods to improve overall accuracy assuming a fixed existing underground loop detector infrastructure. Very few studies are conducted location placement spacing detectors for urban expressways which vital when potential specific road section be monitored. Guidelines regarding this can highly beneficiary developing proactive safety management systems under budget constrains hazardous sections. This study evaluated six spacings by separate models with Bayesian Network. Crash data traffic flow (flow speed) were collected two years (December, 2006 November, 2008) from 30 Shinjuku 4 Tokyo Metropolitan Expressway, Japan. aggregated each 250 meters segment (in case, one upstream other downstream consideration) evaluated. 5-minute cumulative difference average speed between down stream found suitable predictors. Although all could predict more than 50% future crashes accurately, placed 500 center meter provided 63% 80% non-crash situations.