Multi-state and multi-sensor incident detection systems for arterial streets

Incident detection systems typically emphasize incident presence and locati on over incident severity and incident recovery. Yet, Advanced Traveller In formation Systems(ATIS) and Advanced Traffic Management Systems (ATMS) rely on the latter states to implement and terminate diversion, and its support ive control strategies. Further, incident detection systems directly benefi t from processing measurement vectors rather than scalars. Vectors of lane measurements favor detection through lane imbalances and identification of incident host lanes. Intelligent Transportation Systems promise new sensor data to control centers, including the travel times experienced by probe ve hicles. Vectors of new and old sensor inputs may possess enhanced discrimin atory values. To accomodate added detection states and the fusion of multi-sensor input v ectors, this paper reformulates the arterial incident detection problem as a multiple attribute decision making problem with Bayesian scores. This nov el approach utilizes as input the combinations of simulated probe travel ti mes, number of pn,be reports, lane specific detector occupancies and vehicl e counts. Models based solely on probe data lack in performance due to exce ssive overlaps in class distributions. Models based on detector occupancies and vehicle counts by lane perform outstandingly. They display a propensit y to detect through lane measurement imbalances. The probe data is shown to enhance the performance of detector data based models. (C) 1999 Published by Elsevier Science Ltd. All rights reserved.