走廊层面交通小区划分方法的优化

基于公交客运走廊划分的交通小区，可为走廊公交线网优化提供更为符合公交乘客出行特征的需求分析和预测基础.本研究针对公交客运走廊特点，将走廊划分为直接影响区和间接影响区，提出分层次的交通小区划分理论.针对走廊不同影响区在需求预测和分析中对结果精度要求的不同，基于大数据选取更加适合公交走廊交通小区划分的聚类指标，提出直接影响区细分和间接影响区合并的划分方法，并通过引入聚类因子初步确定聚类数目和交通小区中心对传统聚类方法进行改进，克服了传统聚类方法随机选取聚类数目和中心而影响聚类精度的不足.最后，基于多源异构大数据对广渠路走廊进行实例验证，结果表明，本研究提出的分层次小区划分方法较传统方法在适用于公交客运走廊需求分析和预测划分精度方面更优.

Optimization Method of Traffic Analysis Zones Division in Public Transit Corridor

Defining traffic analysis zone (TAZ) based on public transit corridor provides a demand analysis and forecast basis for the bus line optimizations in the corridor which also accounts for bus passenger characteristics for analysis. This paper proposed a hierarchical division method for public transit corridors, which divided areas along the corridor as directly and indirectly influenced TAZs. Considering different requirements for the results' accuracy in the directly and indirectly influenced TAZs, this paper selected clustering indicators that are more suitable for the TAZ division in public transit corridors based on big data. It then proposed the different clustering methods for the directly and indirectly influenced TAZs. The paper also introduced the clustering factor which can initially determine the number of clusters and the center of the traffic area. To illustrate the applicability of the proposed method, this paper presented a case study using the big data from the public transit corridor of Guangqu Road in Beijing, China. The results indicate the TAZ division based on public transit performs better than the traditional method