融合GPS轨迹和POI数据关联规则的城市功能区识别  

Identification of Urban Functional Areas Based on GPS Trajectory and POI Data with Association Rules

作　　者：冯慧芳 杨文亮 FENG Huifang;YANG Wenliang(College of Mathematics and Statistics,Northwest Normal University,Lanzhou 730070,China)

机构地区：[1]西北师范大学数学与统计学院,甘肃兰州730070

出　　处：《测绘科学技术学报》2020年第4期414-420,共7页Journal of Geomatics Science and Technology

基　　金：国家自然科学基金项目(71761031,71561024)。

摘　　要：伴随城镇化的深入,城市化进程逐渐加快,城市空间结构不断变化。识别城市功能区及其分布特征,对于优化城市空间结构具有重要意义。动态城市大数据的涌现为城市空间结构的研究提供了新视角。以城市出租车GPS轨迹数据和城市POI数据为基础,提出了一种基于Apriori关联规则算法的功能区识别算法,并对兰州市城市功能区进行了实证研究。首先,结合出租车GPS轨迹和POI数据,应用Apriori关联规则算法提取强关联规则;然后,结合城市栅格方法,构建栅格关联规则矩阵和功能区域识别指标,提出了功能区识别算法;最后,实证研究兰州市城市功能区,并结合富集因子、基于核密度估计的交通热点区域挖掘验证功能区域识别算法的有效性。With the deepening of urbanization,the urbanization process is gradually accelerating.The urban spatial structure is constantly changing.It is particularly important to identify urban functional areas and their distribution characteristics for optimizing urban spatial structure.The emergence of dynamic urban big data provides a new perspective for the study of urban spatial structure.Based on the taxi GPS trajectory and POI data,an identification algorithm of urban functional areas using Apriori algorithm is proposed in the paper.An empirical study on urban functional area of Lanzhou is studied.Firstly,the strong association rules are extracted based on Apriori algorithm by combining taxi GPS trajectory and POI data.Then,combined with the urban grid method,the grid strong association rule matrix and the functional area identification indexes are constructed,and an identification algorithm of urban functional areas is proposed.Finally,an empirical study on urban functional area of Lanzhou is studied.The effectiveness of the identification algorithm is verified using enrichment factor and mining the traffic hotspot areas based on kernel density clustering.

关 键 词：城市功能区 关联规则 核密度估计 GPS轨迹 POI数据 

分 类 号：P208[天文地球—地图制图学与地理信息工程] TP393[天文地球—测绘科学与技术]