基于潜在类分析的城市异质出行群体识别研究

Identification of heterogeneous urban traveler groups based on latent class cluster analysis

异质出行群体的合理划分是提升出行需求预测准确性和实施主动式交通管理的关键,针对目前城市多方式出行群体划分研究的不足,在分析出行习惯与偏好差异影响因素的基础上,提出基于潜在类分析(Latent Class Cluster Analysis,LCCA)的城市异质出行群体识别方法.以北京市为例,应用揭示性偏好调查进行基础数据收集,运用Mplus软件编程实现LCCA模型估计.模型将出行者划分为三类异质出行群体,群体1:低出行+方式均衡组(20.4%),群体2:中高出行+小汽车偏好组(30.3%),群体3:高出行+绿色交通组(49.3%).模型回归结果表明:群体2、3的百分比与北京市小汽车、公共交通出行比例之差均不超过2%,证明提出的出行群体识别方法有效,个人属性、出行者对各交通方式的认知与态度对群体隶属影响显著.针对各异质出行群体提出了相应的绿色交通发展措施,为城市交管部门的精细化出行管控提供重要依据.

The reasonable division of heterogeneous traveler groups is key to improving prediction accuracy for travel demand forecast and implementing effective active traffic management.To address the lack of studies on traveler group division in urban multimode transportation,based on influencing factors analysis of travel habits and preference differences,a research methodology of heterogeneous traveler group identification based on Latent Class Cluster Analysis(LCCA)is proposed.Beijing is chosen as the place to conduct a case study,and the Revealed Preference(RP)survey is applied to collect the basic data.The LCCA model is estimated by programming in Mplus.The model identifies three heterogeneous traveler groups:Group 1(20.4%)travel less fre-quently and prefer both the automobile and green modes.Group 2(30.3%)travel in moderate frequency and prefer the automobile.Group 3(49.3%)travel more frequently and prefer to travel by green modes.The results of the regression model show that the difference between the percentage of Group 2 and 3 and that of trips by automobile and public transport in Beijing is no more than 2%,proving that the proposed traveler group identification method is effective.Individual characteristics,travelers’cognition of and attitude towards different travel modes have significant impacts on group affiliation.In light of the findings,measures to promote green transportation development for heterogeneous traveler groups are put forward,which lays an important basis for implementing refined management aimed at different traveler groups.