基于众包数据的城市慢行空间关联网络推荐  

A Recommended Linking Network Study of Urban Walking Space Based on Volunteered Geographic Information(VGI)Data

作　　者：张乔嘉 赵渺希[1] 师浩辰 ZHANG Qiaojia;ZHAO Miaoxi;SHI Haochen(State Key Laboratory of Subtropical Building and Urban Science,School of Architecture,South China University of Technology,Guangzhou 510640,China)

机构地区：[1]华南理工大学建筑学院亚热带建筑与城市科学全国重点实验室

出　　处：《城市与区域规划研究》2024年第2期130-147,共18页Journal of Urban and Regional Planning

基　　金：国家社科基金重大项目“新型基础设施推动粤港澳大湾区经济发展与机制运行一体化研究”(21&ZD107)资助。

摘　　要：准确识别、合理划分游憩停留区并梳理它们之间的关联网络,对于提升市民游憩体验和提高旅游设施要素配置具有重要意义。文章提出基于众包GPS轨迹数据探究慢行关联网络推荐效应的新路径。文章以广州老城为案例开展应用研究,结果显示:老城区游憩停留热点的空间分布和轨迹交互特征明显;在三种关联网络的推荐算法比较中,尽管区位推荐算法(LRM)在复现二模网络“轨迹-停留热点”删失数据的精度有限,但在推荐视角下的资源非对称分配算法中能生成较多的链式结构,为城市慢行空间网络延长游览路径、降低集聚比例、分散风险提供了可能,为优化广州老城慢行空间的关联网络提供了有效的规划支撑。Accurately identifying and reasonably dividing recreational stay areas and analyzing their connections are of great significance for improving citizens’recreational experience and optimizing the allocation of tourism facilities.The paper proposes a new approach to explore the recommended effect of slowing-moving association networks based on the V-GPS(volunteered GPS)data.Taking an old town in Guangzhou as a case,this paper conducts a practical study.The results show that the spatial distribution and trajectory interaction characteristics of recreational hotspots in the old city area are obvious;in the comparison of recommended algorithms for the three types of associated networks,although the location recommendation algorithms(LRM)has limited accuracy in reproducing the deleted data of the second mock examination network,the asymmetric resource allocation algorithm from the recommended perspective can generate more chain structure,which makes it possible for the urban walking space network to extend its touring paths,avoid the over-concentrated recommendations and disperse risks,providing effective planning support for improving the association network of the urban walking space of the old city in Guangzhou.

关 键 词：慢行空间 众包GPS轨迹 停留区 关联网络 

分 类 号：TU9[建筑科学]