Research and Application on Highway Traffic Index System
Yuan Zheng,Bin Ran,Zhengfeng Lu,Xu Qu,Jian Zhang
DOI: https://doi.org/10.1061/9780784481523.147
2018-01-01
Abstract:To evaluate highway traffic operation conditions, this paper researches and applies the highway traffic index. Data were used from the I-880N highway with location 401240 in California, United States. Cross correlation analysis was applied to divide the highway traffic state. Then k-means clustering and quest decision tree model were used to determine the specific value of highway traffic index. Results show that highway traffic index is divided into five levels: smooth, basic smooth, mild congestion, moderate congestion, and severe congestion. The five levels are quantified as (0-2), (2-4), (4-6), (6-8), and (8-10); the division threshold of traffic density at each level is 21.93, 34.82, 41.09, 58.00, and 75.35pcu/mile/lane respectively; color classifications are green, light green, light yellow, yellow, and red; real-time release and release cycle selected based on actual situation. This study establishes a highway traffic index system and provides a theoretical basis for follow-up application.