A new method for incident detection on urban arterial roads

This paper presents a Bayesian network based incident detection algorithm for urban arterial roads. The algorithm is capable of detecting lane-blocking incidents whose effects are manifested by patterns of deterioration in traffic conditions. A new detector configuration is proposed to collect lane volume and occupancy data. The traffic signal scheduling is incorporated into incident detection algorithm. A dynamic Bayesian network is constructed and trained to detect incidents. Off-line tests using simulated data are conducted to assess the performance of the algorithm. The testing results are encouraging, and Bayesian network based approach is considered promising.