车辆轨迹信息的道路属性粗糙集定权模型

A rough set weight model of road set theory attribute based on vehicle trajectory information

针对利用道路自身属性来计算重要度,并进行道路选取时产生的路网不连通、道路悬挂等问题,该文提出一种新的道路属性定权模型。结合轨迹大数据与道路自身属性,以路段为基本单位,选取道路长度、连通度、网络密度、车流量、平均车速、偏好系数6项属性作为道路重要性的衡量指标,采用基于牵引程度的离散化方法,实现指标数据的离散化,利用粗糙集确定属性权重。基于武汉市车辆的轨迹数据和道路数据的实验结果表明,该文模型计算得到的重要性值更符合实际情况,以该重要性进行道路选取时,有效地解决了路网中一些道路由于自身属性较小,使得重要度计算较小,导致道路选取时被舍弃,产生的路网不连通、道路悬挂等问题。

In this paper,a new weighting model of road attributes was proposed,which could be used to reduce road suspension and road network disconnection while the importance degree of roads is used to select roads.Based on the trajectory data and the road’s own attributes,this method took each road segment as one basic unit.Six attributes such as road length,connectivity,network density,traffic flow,mean velocity,and preference coefficient were selected as indicators of basic unit’s importance.Then each indicator data was discretized by traction degree.Finally,each attribute weight was determined by rough set.The experiment was carried out on the trajectory data and road data of Wuhan city.And the results showed that the importance of the road calculated by the model better reflected the actual situation.When some roads of less importance were deleted,the attendant problem such as road suspension and road network disconnection could be solved.