Abstract: The park-and-ride mode of commuter corridor based on rail transit is an effective means to improve the corridor travel rate ，alleviate the urban traffic congestion and solve the insufficiency of parking facilities.Based on structural equation model (SEM)， the personal attributes，the information demand preference and the park-and-ride facilities level were taken as exogenous latent variables; the travel purpose，the departure time，the travel mode and the travel intensity were taken as endogenous latent variable; the P&R travel behavior path and the effectiveness of various factors were analyzed.Based on the actual survey data，LISREL software was used to evaluate the effect of the model.The analysis results show that：①The travel mode of P&R has strong regularity in time， and travelers have a higher percentage of choosing P&R travel mode on morning and evening peak hours.②Traffic state information， accident information，parking information have significant effect on P&R travel behavior.Among them，the traffic state utility value is 0.8; in the aspect of endogenous latent variables，the travel purpose has the highest priority，which has greater impact on the departure time，travel mode and travel intensity.

Key words: traffic engineering, park & ride, structural equation model (SEM), travel behavior mechanism

摘要： 通勤廊道上基于轨道交通的停车换乘出行模式是提高廊道出行率，缓解城市交通拥堵、解决停车设施不足的有效手段。基于结构方程模型(SEM)， 选取个人属性、信息需求偏好及停车换乘设施水平作为外源潜变量，选择出行目的、出发时刻、出行方式及出行强度作为内生潜变量，分析了停车换乘 (P&R)出行行为产生路径及各因素的效用。以实际调查数据为基础，利用LISREL软件，对模型效果进行评价。结果表明：① P&R出行方式在时间上具有较 强规律，早、晚高峰时段选择P&R出行方式的比例较大；② 路况状态信息、事故信息、停车信息对P&R出行行为影响显著。其中，路况状态的效用值达到 0.8；内生潜变量方面，出行目的具有最高的优先级别，其对出发时刻、出行方式及出行强度的影响较大。

关键词: 交通工程, 停车换乘(P&R), SEM模型, 出行行为机理