信息影响下区分潜在类别的公交线路选择模型

Bus route choice model under influence of information accounting for latent class

为分析实时公交运行信息和用户自身偏好对公交用户路径选择行为的影响,引入三角分布描述公交信息的不确定性,并使用前景函数替代效用函数描述出行者决策的不完全理性,建立基于前景函数的选择概率模型。为进一步区分不同出行者偏好差异,在前景模型的基础上结合潜在类别模型,构建区分出行者潜在类别的前景概率模型。利用2013年北京市的实际调查数据进行参数估计并检验。结果表明:a)出行者的选择行为受自身的偏好影响;b)基于景函数的选择模型比传统的效用函数模型拟合度更好、更贴合实际选择行为;c)区分潜在类别模型的统计学特征更优秀;d)多数公交用户对等车时间更敏感。通过各种措施缩短等车时间或直觉等车时间比提高公交运行速度更能吸引乘客。

In order to analyze the effect of real time bus operation information and bus travelers＇ preference on route choice behavior, this paper described the uncertainty of bus information by using triangular distribution, replaced the utility function with prospect theory function to represent the incomplete rational decision of bus travelers, and established the choice probabili- ty model based on the prospect theory. For the purpose of classifying the different travelers＇ individual preference heterogenei- ty,it established a prospect probability model of distinguishing latent class by integrating the latent class model with the pros- pect model,then it estimated the model＇ s parameters and tested the model by utilizing the actual survey data of Beijing （2013）. The results indicate that： a）Bus travelers＇individual preference has an influence on their route choice behavior; b） The route choice model based on prospect theory has a better goodness-of-fit and is more realistic with the actual choice behavior than the traditional utility function model;c ）The prospect theory selection probability model has better demographic characteristics; d） Most bus travelers are more sensitive to waiting time. Cutting waiting time or instinctive waiting time is more attractive than improving bus operation velocity.