详情

原文传递 EVALUATION OF ONLINE TRAVEL TIME ESTIMATORS AND PREDICTORS.

题名：	EVALUATION OF ONLINE TRAVEL TIME ESTIMATORS AND PREDICTORS.
作者：	Lindveld-CDR; Thijs-R; Bovy-PHL; Van-der-Zijpp-NJ
关键词：	Accuracy-; Dynamic-traffic-management; Estimating-; Field-tests; France-; Italy-; Loop-detectors; Netherlands-; Online-estimating; Traffic-congestion; Travel-time
摘要：	Travel time is an important characteristic of traffic conditions in a road network. Up-to-date travel time information is important in dynamic traffic management. Presented are the findings of a recently completed research and evaluation program called DACCORD, regarding the evaluation of tools for online estimation and prediction of travel times by using induction loop detector data. Many methods exist with which to estimate and predict travel time by using induction loop data. Several of these methods were implemented and evaluated in three test sites in France, Italy, and the Netherlands. Both cross-tool and cross-site evaluations have been carried out. Travel time estimators based on induction loop detectors were evaluated against observed travel times and were seen to be reasonably accurate (10% to 15% root mean square error proportional) across different sites for uncongested to lightly congested traffic conditions. The evaluation period varied by site from 4 to 30 days. Results were seen to diverge at higher congestion levels: at one test site, congestion levels were seen to have a strong negative impact on estimation accuracy; at another test site, accuracy was maintained even in congested conditions.
总页数：	Transportation Research Record. 2000. (1719) pp45-53 (4 Fig., 4 Tab., 10 Ref.)
报告类型：	科技报告