基于改进PageRank算法的路网重要交叉口筛选方法

Dynamic Sorting Method for Road Network Primary Intersections Based on PageRank Algorithm

为便于对饱和交通状况下的城市道路交叉口进行分级管理,需解决城市道路交叉口的重要性排序问题,综合考虑全路网中各交叉口之间的静态结构连接关系和动态流量影响,在改进PageRank算法的基础上,提出了能够反应全路网动态变化的交叉口繁忙程度指标,并将该指标用于路网重要交叉口排序筛选来分析交叉口的状态.研究结果表明:排序越靠前的交叉口越繁忙也越重要,交叉口繁忙程度指标综合考虑了全路网交叉口状况,弥补了以饱和度为评价指标只能片面衡量单个孤立交叉口状态的不足,更准确地反映了饱和交通状况下交叉口之间的相互影响;本文方法排序结果与饱和度评价指标排序结果相比,40%交叉口的排序升降幅度在3位以内,30%交叉口的排序平均下降了7位,其余30%交叉口的排序平均上升了8位.该研究结果为饱和交通状况下交叉口的合理分级提供了量化手段,有助于及时发现急需管控的交叉口.

In order to classify intersections of a city under saturated traffic operating conditions, sort- by-importance measures for intersections should be provided. Considering the static connections and dynamic influences of traffic volume between intersections, a key busyness index that can reflect the busy degree of each intersection in the road network is proposed on the basis of a modified PageRank algorithm. This busyness index is then used for sorting important intersections in the road network and analyzing their situations. Results show that the intersections that are sorted in the front are very busy and important. By taking into consideration of the whole intersections in a road network, the busyness index can make up the disadvantage of the saturation index that can only evaluate the situation of single intersections, and therefore is capable of reflecting accurately the mutual influences of the connection relations among intersections under saturated conditions. Compared with the sorted results by saturation index, the ranking orders by business index of 40% intersections change within 3 positions, 30% drop by 7 positions, and 30% rise by 8 positions. The obtained findings can be used to classify intersections of a city under saturated traffic conditions to timely find the key intersections.