文章编号： 1672-5328（2022）06-0056-09

翁剑成1，潘晓芳1，邸小建2，林鹏飞3
(1. 北京工业大学交通工程北京市重点实验室，北京100124；2. 交通运输部规划研究院，北京100028；3. 北京工 业大学信息学部，北京100124)

摘要： 公共汽车满意度反映了乘客对公共汽车服务质量的直接感受，现有研究缺乏对不同乘客群体 满意度差异以及公共汽车运行质量与满意度之间关系的研究。首先构建公共汽车满意度评价指标体 系，包含时效性、安全性、便捷性、舒适性、可靠性、经济性6 个一级指标，21 个二级指标和77 个 三级指标。通过问卷调查得到北京市100 条公共汽车线路共3 012 份满意度调查数据，利用高斯混 合模型聚类方法识别出4 类乘客群体，应用随机森林特征选择方法鉴别不同群体满意度的核心偏好 因素。将满意度调查数据与公共汽车平均运行速度、满载率、到站时间波动性等数据进行关联匹 配，构建针对工作日高峰时段的公共汽车满意度与运行质量交互影响模型，定量评估公共汽车运行 状态变化对满意度的影响。通过实例验证了影响模型的准确性，继而提出基于不同乘客群体的公共 汽车满意度提升策略。

关键词： 公共交通；公共汽车；满意度；群体差异；评价指标；服务质量；运行质量

中图分类号： U491.1+7

文献标识码：A

Strategies for Improving Bus Transit Satisfaction Based on Differences in Passenger Groups

WENG Jiancheng1, PAN Xiaofang1, DI Xiaojian2, LIN Pengfei3
(1. Beijing Key Lab of Traffic Engineering, Beijing University of Technology, Beijing 100124, China; 2. Transport Planning and Research Institute Ministry of Transport, Beijing 100028, China; 3. Faculty of Information Technology, Beijing University of Technology, Beijing 100124, China)

Abstract： Bus transit satisfaction reflects the quality of bus services perceived by passengers. However, current studies are limited on the difference of satisfaction among various passenger groups and the relationship between bus operational quality and satisfaction. This paper starts with the development of an assessment indicator system for bus transit satisfaction that identifies six first-level indicators, including timeliness, safety, convenience, comfort, reliability, and economic efficiency, as well as 21 second- level and 77 third-level indicators. Data from 3,012 satisfaction surveys covering 100 bus lines in Beijing are obtained via questionnaires. Four types of passenger groups are selected using a clustering method of Gaussian Mixture Model. Furthermore, the core preference factors for satisfaction among different passenger groups are identified using the Random Forest feature selection method. By correlating the satisfaction survey data with bus transit data, such as average bus operation speed, load factor, and arrival time fluctuation, the paper develops an interactive impact model between bus transit satisfaction and operational quality during peak hours on workdays. Meanwhile, the paper quantitatively assesses the impact of bus operation status changes on the satisfaction of different passenger groups. Finally, after verifying the accuracy of the interactive impact model with examples, the paper proposes strategies for improving the bus transit satisfaction of different passenger groups.

Keywords： public transit; bus transit; satisfaction; group difference; assessment indicators; service quality; operational quality