详情

原文传递 Integration of Point-Based and Link-Based Incident Detection and Traffic Estimation.

题名：	Integration of Point-Based and Link-Based Incident Detection and Traffic Estimation.
作者：	Mahmassani-H.S.; Haas-C.; Logman-H.; Shin-H.J.
关键词：	Traffic-surveillance; Automatic-vehicle-identification.;Speed-estimation; Monte-Carlo-modeling; Modeling-sensors; Recommendations-; Information-flow-architecture; Link-based-; Point-based-; Travel-time; Traffic-speed.
摘要：	The implementation of Intelligent Transportation Systems (ITS) carries the promise of more efficient use of already existing transportation networks. Reliable speed estimation and timely incident detection reduces congestion on highways. The Inductive Loops Detectors (ILD) and Automated Vehicle Identification (AVI) are the main sensing systems deployed by many traffic management agencies to collect speed and travel time data. They represent point and link type data, respectively. This study utilized the Bayesian Updating and Weighted Average methods to estimate highway speed and travel time by integrating the AVI and ILD data extracted from San Antonios Traffic Management Center (TransGuide). The analysis considered the sensors accuracy, reliability, and level of penetration. The results of the analysis support the reliability of the AVI system for speed and travel time estimation and the ILD system for occupancy, point-based speed measurement and for Automatic Incident Detection (AID) algorithm processing. Additionally, Monte Carlo simulation model was designed to model sensors fusion to detect traffic incidents. The Monte Carlo model showed promising results when validated using traffic and incident data from the San Antonio network. It could be used as a performance predictor that supports traffic sensing systems investment decisions.
报告类型：	科技报告