关键词: 交通小区, 空间关联矩阵, 空间聚类分析, 自动划分

Abstract: Traffic zone is an advantageous tool for analyzing complex transport network of urban.It has similar traffic characteristics and strong traffic relationship within one traffic zone.Thus，it can simplify the complexity of the network.But it still rests on application about this research around the world presently.Based on analyzing concept and properties of traffic zone，a method of spatial statistic analysis is proposed to research automated division of traffic zone.Model of spatial clustering analysis algorithm is offered，and parameters models are built，such as spatial relational matrix，traffic correlation.The flow and algorithm are designed for traffic zone division.Finally，combined with floating car data and actual transport network in Shanghai，the automated division of traffic zone is realized and it is concluded.

Key words: traffic zone, spatial relational matrix, spatial clustering analysis, automated division