Abstract: SP survey is very popular in transportation research in countries except of China now. This paper gives a general description of the different concepts of RP and SP survey firstly. Due to the relativity between RP survey data, it is very difficult to calibrate the value of time in mode split model. This paper takes Beijing SP survey for an example, analyses the reason why the value of time is not easy to calibrate, and adopts SP survey to get enough data. Finally, the authors accurately calibrate parameters of the value of time and the accuracy of the model has been improved.

Key words: SP survey, traffic model, mode split, behavior value of time

摘要： 国外已经广泛使用的SP调查方法在我国交通领域还鲜有应用，因此，本文首先介绍了交通SP（Stated Preference）调查和RP（Revealed Preference）调查的不同，并阐述了前者相对于后者的优越性。交通方式选择模型中，采用RP调查标定时间价值参数，通常会由于数据相关性较强而产生误差，进而影响到整个四阶段模型的精度。本文在构架交通方式选择模型的基础上，指出行为时间价值参数的重要性，并分析了其不易准确衡量的原因。在此基础上，以北京市市域交通模型为例，利用SP调查以较低的成本获取足够的数据，从而准确地标定了时间价值参数，提高了模型的精度。

关键词: SP调查 , 交通模型 , 交通方式选择 , 行为时间价值