Abstract: Value of drivers’ commuting travel time is an important parameter when implementing and appraising a road congestion pricing project. Both Logit model and Mixed Logit model whose time parameter is fixed and cost parameter has a lognormal distribution are used to estimate the value of drivers’ commuting travel time by using SP data from drivers in Dalian city. The result indicates: the value of t-statistic are significant in both Logit model and Mixed Logit model; Mixed Logit model estimation has a better value of goodness-of-fit index and maximum likelihood than Logit model; Value of drivers’ commuting travel time is 24.58RMB/h in Logit model estimation; Value of drivers’ commuting travel time has a lognormal distribution whose parameters are 2.73 and 1.09 in Mixed Logit model. The value of drivers’ commuting travel time from Mixed Logit model estimation is recommended when implementing and appraising a road congestion pricing project.

Key words: value of travel time, stated preference survey, Logit model, Mixed Logit model

摘要： 私家车出行者的通勤出行时间价值是实施和评价道路拥挤收费项目的重要参数之一。本文利用大连市私家车出行者意向调查数据，分别使用Logit模型和时间系数为常数、费用系数服从对数正态分布的Mixed Logit模型来标定私家车出行者通勤出行时间价值。标定的结果显示：两个模型标定结果中的各参数的t值都非常显著；Mixed Logit模型标定结果中的最大似然值和优度比都优于Logit模型；使用Logit模型标定的私家车出行者通勤出行时间价值为24.58元/小时；使用Mixed Logit模型标定的私家车出行者通勤出行时间价值服从参数为2.73和1.09的对数正态分布。在实施和评价道路拥挤收费项目时，推荐使用Mixed Logit模型的标定结果。

关键词: 出行时间价值, 意向调查, Logit模型, Mixed Logit模型