Abstract:

This study aims to investigate the heterogeneous perception of travelers on urban traffic states. First, a video- based experiment was conducted for collecting the perception data. Then, an order Logit Bayesian hierarchical model was developed to estimate the individual- level perception parameters. The heterogeneity of travels’perception on traffic state can be investigated through examining the individual parameters by accounting for the effects of objective road- and traffic-related factors. The results show that: (1) Both road attributes (number of lanes or road segment width) and traffic flow factors (traffic volume and average travel speed) have significant effects on traveler’s perception. (2) The effects of these factors significantly vary among participants. The variations in effect of road attribute factors are more significant than those of traffic flow factors. Average travel speed has a relatively stable effect on their perceptions. (3) The difference in personal attributes is one of the main hints for the heterogeneous perception.

Key words: traffic engineering, traffic state identification, individual perception heterogeneity, Bayesian hierarchical order Logit model

摘要：

为研究出行者在感知城市道路拥堵状态时的个体差异性，首先，选取长沙市 5条典型路段，采用视频观测法，设计拥堵状态主观感知实验，收集出行个体对城市道路拥堵状态的真实感知数据；其次，基于多层有序 Logit贝叶斯模型，获取出行者个体感知参数，深入分析道路及交通流因素对出行者拥堵感知影响的差异程度，并进一步探究造成感知差异的个体属性因素.结果表明：所选道路和交通流特征因素，对出行者的拥堵感知结果有显著影响；在其他变量不变的情况下，道路特征因素对不同出行者拥堵感知结果的影响差异性大于交通流特征因素，其中平均车速的影响差异性最小；个人属性的不同，是造成感知差异的主要原因.

关键词: 交通工程, 交通拥堵感知, 个体感知差异, 多层有序Logit贝叶斯模型