Researches on the Calculation Method of Saturation Flow Rate at Signalized Intersection Based on K-means Clustering Algorithm

摘要

This paper studies K-means clustering algorithm for the calculation of saturation flow rate at signalized intersection. The calculation results are based on the measured traffic data, which are surveyed at the intersection of Xinhua Road and Youyi Road in Tangshan. By analyzing the distribution characteristics of average headway in a day, this paper presents the negative influences of unsaturated or oversaturated traffic flow on headway. A portion of the average headway which affected by unsaturated or oversaturated condition are removed by K-means clustering algorithm. The developed method is used to calculate the saturation flow rate within different conditions. The calculation results illustrate that the proposed method can offset the effects of unsaturated or oversaturated condition.