Statistical and neural classifiers to detect traffic operational problems on urban arterials

✍ Scribed by Sarosh I. Khan; Stephen G. Ritchie

• Book ID: 104368917
• Publisher: Elsevier Science
• Year: 1998
• Tongue: English
• Weight: 345 KB
• Volume: 6
• Category: Article
• ISSN: 0968-090X
• DOI: 10.1016/s0968-090x(99)00005-4

✦ Synopsis

This paper proposes using arti®cial neural networks in a modular architecture to assist in detecting dierent types of operational problems on signalized urban arterials. A trac surveillance infrastructure that includes inductive loop detectors on intersection approaches as well as mid-block system loops for trac monitoring is used. For arterials, problems that require the attention of a trac management center operator include lane-blocking incidents, special event conditions, and detector malfunctioning. Problem detection depends on factors such as operating conditions, con®guration of sensors within the network, and block or link length. The feasibility of training and testing neural network models as components of a modular architecture, with an appropriate model for each sub-problem of pattern recognition, is demonstrated. The performance of this modular architecture exceeded that of any single architecture applied to the detection of dierent types of operational problems. The paper also reports on the performance of each type of model considered and the eect trac ¯ow levels and detector con®guration have on the performance of the incident detection model. The results show that with the selection of a suitable architecture, the modular neural classi®er outperforms alternative discriminant analysis-based classi®ers. This is demonstrated using cyclic data from microscopic simulation and ®eld data from Urban Trac Control System (UTCS) implementations in the Cities of Los Angeles and Anaheim, California.