中文摘要：

从公交线路客流调查样本数据出发,拟合客流到达曲线,将中间站客流特征转化到首站.基于客流曲线特征,建立了公交时段的聚类模型,该模型综合考虑类内部的相似性和类间的差异性.基于前2部分成果,建立公交发车间隔模型,该模型综合考虑乘客时间成本和车辆的载客率.以常州市公交线路B22的跟车调查数据为基础,进行实证研究,得出公交发车间隔和发车时刻表,验证了模型的可行性和实用性.

英文摘要：

This paper conducts a study on bus departure timetable. At first, the curve of passenger-ar- rival rate is fitted based on survey data of bus stations; then, a clustering model of transit operating periods by the changes of arriving station passengers is raised, which takes into account the similarity within class and the differences between classes; finally, according to clustering results of above mod- el, bus departure interval model is established, which consider synthetically time value of passengers and bus load rate. One bus line B22 survey date of Changzhou City is exploited to verify the method and models in every section and obtain satisfactory results, so the method and models is feasible and practical.