居民出行调查中交通小区划分方法的改进  

Improvement of the Traffic District Partition in Resident Trip Investigation

作　　者：赵锦焕[1] 李文权[1] 

机构地区：[1]东南大学,交通学院,南京210096

出　　处：《交通运输工程与信息学报》2009年第2期110-115,共6页Journal of Transportation Engineering and Information

基　　金：国家高科技研究发展计划项目(2007AA11Z200)

摘　　要：为了减少居民出行调查的工作量和费用,利用以往调查内容中的年龄、职业、收入和车辆拥有率等指标,根据聚类分析的等价矩阵模糊聚类方法,通过MATLAB编程进行聚类分析,用F-统计量确定最佳模糊聚类类数,对交通小区重新分类。在调查时,只需在每类小区选择一个交通小区,就可满足调查精度要求。最后用模糊C-均值聚类方法验证。计算结果表明,模糊聚类分析方法可以应用在居民出行调查的交通小区划分中。In order to reduce the workload and cost in resident trip investigation, data including age, profession, incomes and vehicle ownership rate of the late survey were used. According to the equivalent matrix fuzzy of cluster analysis, and with MATLAB programming, the optimal classification numbers were determined using F-statistics, and the traffic districts were re-classified. By this way, only a traffic district was chosen from each type traffic zone, the accuracy will meet the demands when conduct a traffic investigation. To verify the method, a fuzzy C-means cluster was used. The results showed that the fuzzy cluster analysis method can be conveniently applied to the partition of traffic zone in resident trip investigation.

关 键 词：交通小区 模糊聚类 出行调查 类数 

分 类 号：U491.121[交通运输工程—交通运输规划与管理]