卫星城居民出行特征及出行方式选择研究是解决卫星城交通问题的基础和前提。以南宁市 浦口区为例，通过Access 数据库对居民出行OD调查数据进行处理挖掘，从居民的出行频次、出 行目的、出行方式、出行时耗及出行时间分布五方面对卫星城居民出行特征进行研究，总结出浦 口区居民出行特征现状；挖掘并量化影响卫星城居民出行方式选择的关键因素，利用浦口区居民 出行调查数据，构建居民出行方式选择多项Logit 模型，通过SPSS软件进行模型拟合和标定。在 模型拟合结果中，以公交车为例进行拟合结果分析，可得到影响选择公交出行的主要因素——性 别、年龄以及有无公交IC 卡，并通过计算以及模型预测值对模型拟合精度进行校验。结果显示， 除非机动车预测效果较差以外，其余预测误差均在可接受范围内。

Study on residents′ travel characteristics and travel mode choice behaviors is the basis and premise to solve traffic problems of satellite city. Taking Pukou district of Nanjing as an example, the res⁃ idents′ travel OD survey data was processed and excavated by Access. The residents′ travel characteris⁃ tics status was summarized from aspects of travel frequency, travel purpose, travel mode, travel time and its distribution. The major factors influencing the travel mode choice of satellite city residents were exca⁃ vated and quantified. By data mining, the multinomial travel choice Logit model was established. It was fitted and calibrated through SPSS. Taking the public transport as an example, the major factors influenc⁃ ing the choice of bus travel which were gender, age and whether having bus IC card were conducted. The fitting accuracy was tested using the calculation and predicted value. The result shows that the errors of predicted values are acceptable except for non-motorized vehicle.