Timely and accurate incident detection using automated incident detection(AID) systems is considered essential to effectively tackle incident induced congestion problems and to improve traffic management. The core of an AID system is the incident detection algorithm which interprets real time traffic data and makes decisions on incidents. The AID algorithms discussed in this study aim to detect lane-blocking incidents when their effects are manifested by certain types of deterioration in traffic conditions. Theperformance of AID algorithms is normally evaluated against three measures: detection rate, false alarm rate and mean time to detect. In this studyincident detection is treated as a decision making problem rather than aspattern recognition. The focus of AID algorithm design is shifted from precise incident pattern description and reduced traffic pattern misclassification to effective traffic knowledge representation and strong evidentialreasoning capability of the algorithm. Separate algorithms for incident detection on freeways and urban arterial roads are described. For the covering abstract see ITRD E144727. Reprinted by permission of Elsevier.
Samenvatting