基于路网拓扑的聚类分析算法研究与实现

Research and Implementation of Clustering Analysis Algorithm Based on Road Network Topology

聚类分析是一种统计分析方法,其目的是在相似的基础上对数据进行分类。该研究在基于路网拓扑的条件下,提出一种全新的算法实现交通领域中兴趣点集的聚类分析。算法为从集合中任取一个兴趣点,向邻接结点扩展,沿着路网拓扑的方向进行广度搜索,以凝聚度(α)和扩展度(β)为聚类限制条件,将满足限制条件的兴趣点与选定的兴趣点聚类。最后,把算法应用于珠江新城路网,分析行人的可达性情况。

Clustering analysis is a statistical analysis method,which aims to classify the data from their similarities. Based on road network topology,a new algorithm had been put forward to achieve the clustering analysis for points of interest（ POI） in transportation field. The algorithm is that POI taken from point sets would be extend to the adjacent node and has a breadth of search along the direction of the network topology with limiting conditions of aggregating degree（ α） and extending degree（ β）. POI those meet the conditions would be clustered together. At last,this algorithm is applied in the Guangzhou Pearl River Metro network to analysis pedestrian accessibility.