An Analysis of Mode Choice Behavior of Inter-city Travel in Urban Agglomeration Areas Using a Random-parameter Logit Model
-
摘要: 以粤港澳大湾区城市群的广深城际运输通道为例，分析城际运输通道中影响旅客出行方式选择行为的因素及其影响。传统的多项式Logi（tMNL）模型具有无关方案独立性，无法对不同出行者的选择偏好差异进行定量分析，故应用随机系数Logit模型分析城际交通出行选择行为。选取城际出行旅客的个人社会经济属性、心理潜变量（对交通方式舒适性、可靠性和便捷性的心理感受）、城际出行方式特征变量设计问卷。采用线上与线下相结合的方式开展问卷调查，共收集534份问卷，基于此建立并求解随机系数Logit模型。随机系数Logit模型估计结果的伪R2为0.178，表明模型具有良好的拟合度。研究结果表明：城际出行旅客的收入、职业、私家车保有情况、家庭儿童数量以及对出行方式便捷性的感知对其选择行为有显著影响；而出行方式的舒适性、可靠性对城际出行方式的选择行为影响不显著；改善交通方式的便捷性对提升城际出行方式的吸引力起关键作用。因此，在城际交通规划设计、运营管理中应着重考虑便捷性对城际交通方式选择带来的影响。Abstract: In order to study the impact factors of mode choices of travel modes in inter-city transport, the Guangzhou-Shenzhen inter-city transport corridor in the Great Bay Area of Guangdong-Hong Kong-Macao in China is taken as a case study. Given that traditional multinomial Logit models have the issue of independence of irrelevant alternatives (IIA) and they cannot be used to analyze heterogeneous preferences of travelers, a random-parameter Logit model is applied in this study. An online/offline survey is carried out and a total of 534 questionnaires are obtained. In the survey, information including socio-economic attributes of travelers, psychological latent variables (i. e., perceptions toward comfort, reliability, and convenience of travel modes), and attributes of modes for inter-city travel is collected, with which a random-parameter Logit model is developed and estimated. The pseudo R2 of the model is 0.178 at convergence, indicating a good model fit. The estimation results show that the income, occupation, car ownership, number of children in a family, and the perception toward the convenience of travel modes have a significant impact on the mode choice behavior of inter-city travel in urban agglomeration areas, while comfort and reliability of inter-city travel modes are not. Meanwhile, improving the convenience of inter-city travel modes plays a key role in improving their attractiveness. In this sense, special attention should be given to the convenience of the inter-city travel modes in the process of inter-city transport planning and management.
-
表 1 城际出行旅客个人属性
Table 1. Personal attributes of intercity travelers
个人属性 符号 年龄 dage 学历 dschool 月收入 dincome 性别 dfemale 职业 djob 家庭是否有私家车 dcar 是否有小孩 dchild 
下载: 导出CSV
表 2 表征心理潜变量的显示变量
Table 2. Manifest variable represents latent variable
潜变量 显示变量 符号 舒适性
Lcomfort我很难忍受拥挤嘈杂的环境 c1 我很看重交通方式的舒适性 c2 我习惯乘车办公、看书或者娱乐 c3 我遇到交通拥堵、临时停车时会烦躁 c4 频繁的加减速使我感到不适 c5 可靠性
Lreliability我很看重交通方式的准点率 r1 为了更准时到达目的地，我愿意选择舒适性低的交通方式 r2 我不喜欢乘车时发生延误 r3 我出行时会预留充足的时间以防延误 r4 便捷性
Lconvenience我不会为了节省费用而多换乘 v1 我更愿意选择发班频率高的交通方式 v2 我更倾向于上下车在市区的交通方式 v3 我更喜欢乘车流程简单的交通方式 v4
下载: 导出CSV
表 3 情景选择示例
Table 3. An example of SP choice scenario
问卷编号 交通方式 接驳时间/min 乘车时间/min 费用/元 请选择交通方式 A轨道交通 30 53 60 □A 问卷1 B客运班车 20 74 30 □B C私家车 5 60 67 □C A轨道交通 30 80 80 □A 问卷2 B客运班车 20 110 45 □B C私家车 5 90 100 □C A轨道交通 20 80 60 □A 问卷3 B客运班车 13 110 30 □B C私家车 3 90 67 □C A轨道交通 20 53 80 □A 问卷4 B客运班车 13 74 45 □B C私家车 3 60 100 □C
下载: 导出CSV
表 4 个人属性描述性统计
Table 4. Descriptive statistics of personal attributes
个人属性 类别 比例/% ≤25 42.1 年龄/岁 > 25~50 49.5 > 50 8.4 高中及以下 26.3 学历 大专、本科 57.9 硕士及以上 15.9 ≤5 000 60.3 月收入/元 > 5 000~10 000 24.1 > 10 000 15.6 性别 男 54.8 女 45.2 职业 学生 37.2 公务员/事业单位 13.6 企业 23.1 个体户/自由职业/农民/无业 26.1 家庭是否有私家车 有 41.4 无 58.6 是否有小孩 有 42.5 无 57.5
下载: 导出CSV
表 5 模型检验指标结果
Table 5. Fitness statistics of the confirmatory factor analysis
检验指标 RMSEA CFI TLI SRMR 标准 ≤0.08 ≥0.900 ≥0.900 < 0.08 模型参数 0.076 0.92 0.907 0.055
下载: 导出CSV
表 6 验证性因子分析模型结果
Table 6. Results of confirmatory factor analysis
潜在变量 显示变量 标准系数 Z值 p值 Lcomfort c1 0.843 12.778 < 0.001 c2 0.633 13.188 < 0.001 c3 1.157 13.987 < 0.001 c4 0.593 12.053 < 0.001 c5 0.561 11.407 < 0.001 Lreliability r1 0.510 10.924 < 0.001 r2 1.005 14.670 < 0.001 r3 0.270 7.567 < 0.001 r4 0.754 13.933 < 0.001 Lconvenience v1 1.317 15.310 < 0.001 v2 0.420 12.668 < 0.001 v3 0.248 9.875 < 0.001 v4 0.187 8.481 < 0.001
下载: 导出CSV
表 7 不考虑个人属性影响的随机系数Logit回归结果
Table 7. Results of the random-parameter Logit model without demographic characteristics
参数 系数 标准差 z值 p值 非随机项 dbus 0.176 9** 0.065 0.27 0.007 8 drailway 0.467 7** 0.060 0.78 0.004 4 P -0.100 0** 0.003 -3.10 0.000 2 β C -0.007 2 0.003 -1.55 0.120 0 T -0.175 8** 0.005 -3.48 0.000 2 σ Nc 0.000 1 0.015 0.00 0.271 1 NT 0.051 3** 0.012 5.29 0.000 1 注：dbus与drailway表示所选取的城际出行方式是否为客运班车、轨道交通的虚拟变量；Nc与NT表示接驳时间（C）、行程时间（T）的系数的标准差；“**”表示系数在5%的显著性水平下显著。
下载: 导出CSV
表 8 考虑个人属性影响的随机系数Logit回归结果
Table 8. Results of the random-parameter Logit model with demographic characteristics
参数 系数 标准差 z值 p值 T:dage -0.003 0.007 -0.01 0.369 9 T:dschool 0.002 0.088 0.03 0.988 0 T:dincome 0.127** 0.607 2.09 0.003 7 T:dfemale 0.008 0.134 0.62 0.530 2 T:djob -0.051** 0.060 -0.80 0.004 2 T:dcar -0.071** 0.138 -0.52 0.000 6 T:dchild 0.081** 0.142 0.58 0.000 6 T:Lcomfort 0.013 0.014 0.92 0.700 0 T:Lreliability 0.008 0.020 0.38 0.906 2 T:Lconvenience 0.468** 0.031 1.66 0.000 4 注：T：dage表示dage对T的系数的影响，对应式（8）δ3中的1个元素，其余同理。
下载: 导出CSV
表 9 考虑显著的个人属性影响的随机系数Logit回归结果
Table 9. Results of the random-parameter Logit model with significant demographic characteristics
参数 系数 标准差 z值 p值 非随机项 dbus 0.177 2** 0.646 0.27 0.007 1 drailway 0.470 2** 0.601 0.79 0.004 3 P -0.100 3** 0.003 -3.11 0.000 2 C -0.007 1 0.003 -1.58 0.124 4 β T -0.175 7** 0.005 -3.53 0.000 2 T:dincome 0.132** 0.611 2.03 0.004 0 T:djob -0.053** 0.060 -0.80 0.003 9 δ3 T:dcar -0.070** 0.158 -0.52 0.000 6 T:dchild 0.078** 0.140 0.58 0.000 5 T:Lconvenience 0.468** 0.031 1.78 0.000 4 σ NT 0.049 9** 0.012 5.13 0.000 2
下载: 导出CSV
表 10 多项Logit模型回归结果
Table 10. Results of multinomial logit model
参数 系数 标准差 z值 p值 dbus 0.157 0** 0.534 0.51 0.006 0 drailway 0.455 1** 0.575 0.88 0.003 2 P -0.114 3** 0.004 -3.26 0.000 2 C -0.008 2 0.003 -1.58 0.105 6 T -0.187 3** 0.005 -3.66 0.000 2
下载: 导出CSV
-
[1] 向红艳, 任小聪, 陈坚. 城市群内城际短途出行方式选择行为建模[J]. 重庆交通大学学报(自然科学版), 2016, 35(3): 129-133. https://www.cnki.com.cn/Article/CJFDTOTAL-CQJT201603027.htmXIANG H Y, REN X C, CHEN J. Modeling on choice behavior of short-distance intercity travel mode[J]. Journal of Chongqing Jiaotong University(Natural Science), 2016, 35(3): 129-133. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-CQJT201603027.htm [2] 温惠英, 吴亚平, 朱殿臣, 等. 广佛城际交通居民出行特性研究[J]. 重庆交通大学学报(自然科学版), 2017, 36(8): 82-88+94. doi: 10.3969/j.issn.1674-0696.2017.08.15WEN H Y, WU Y P, ZHU D C, et al. Residents'trip characteristics of intercity traffic between Guangzhou and Foshan[J]. Journal of Chongqing Jiaotong University(Natural Science), 2017, 36(8): 82-88+94. (in Chinese) doi: 10.3969/j.issn.1674-0696.2017.08.15 [3] 杨亚璪, 陈芬菲, 陈坚, 等. 基于C-TODIM决策方法的城际出行方式选择模型研究[J]. 铁道运输与经济, 2015, 37(12): 62-68. https://www.cnki.com.cn/Article/CJFDTOTAL-TDYS201512019.htmYANG Y Z, CHEN F F, CHEN J, et al. Study on selection model of intercity traveling modes based on C-TODIM decision-making method[J]. Railway Transport and Economy. 2015, 37(12): 62-68. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-TDYS201512019.htm [4] 林雄斌, 卢源. 都市区跨区域通勤特征与影响因素研究: 以京津城际高铁为例[J]. 城市规划, 2021, 45(12): 104-113. https://www.cnki.com.cn/Article/CJFDTOTAL-CSGH202112012.htmLIN X B, LUY. Intercity commuting in China'smetropolitan area: The case of Beijing-Tianjin intercity high-speed railway[J]. City Planning Review, 2021, 45(12): 104-113. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-CSGH202112012.htm [5] 朱鸿国, 赵文静, 马壮林, 等. 广深运输通道内公共客流分担率预测模型研究[J]. 公路交通科技, 2017, 34(4): 146-153. https://www.cnki.com.cn/Article/CJFDTOTAL-GLJK201704022.htmZHU H G, ZHAO W J, MA Z L, et al. Study on prediction model of share rate of passenger flow inGuangzhou-Shenzhen transport corridor[J]. Journal of Highway and Transportation Research and Development, 2017, 34(4): 146-153. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-GLJK201704022.htm [6] 陈颖雪, 董治, 吴兵, 等. 基于选择方案抽样调查的城市群低频率出行行为研究[J]. 中国公路学报, 2013, 26(3): 158-163. doi: 10.3969/j.issn.1001-7372.2013.03.019CHEN Y X, DONG Z, WU B, et al. Study on low frequency intercity travel behavior of urban agglomeration based on choice-based sampling survey[J]. China Journal of Highway and Transport, 2013, 26(3): 158-163. (in Chinese) doi: 10.3969/j.issn.1001-7372.2013.03.019 [7] 李军, 贾顺平, 钱剑培, 等. 习惯影响下城际出行方式选择意向形成机理[J]. 交通运输系统工程与信息, 2018, 18(2): 7-12, 39. https://www.cnki.com.cn/Article/CJFDTOTAL-YSXT201802002.htmLI J, JIA S P, QIAN J P, et al. Intention formation mechanism in the intercity travel mode choice influenced by the habit[J]. Journal of Transportation Systems Engineering and Information Technology, 2018, 18(2): 7-12, 39. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-YSXT201802002.htm [8] KOUWENHOVEN M, DEJONG G. Value of travel time as a function of comfort[J]. Journal of Choice Modelling, 2018(28): 97-107. [9] 景鹏, 隽志才. 计划行为理论框架下城际出行方式选择分析[J]. 中国科技论文, 2013, 8(11): 1088-1094. doi: 10.3969/j.issn.2095-2783.2013.11.004JING P, JUAN Z C. Analysis of intercity travel mode choice in theory of planned behavior[J]. China Science Paper, 2013, 8(11): 1088-1094. (in Chinese) doi: 10.3969/j.issn.2095-2783.2013.11.004 [10] 吴麟麟, 卢海琴, 汪洋, 等. 引入忠诚度变量的城际出行方式动态选择行为研究[J]. 公路交通科技, 2014(11): 123-129. doi: 10.3969/j.issn.1002-0268.2014.11.020WU L L, LU H Q, WANG Y, et al. Research on intercity travel mode dynamic choice behavior with introduced loyalty variable[J]. Journal of Highway and Transportation Research and Development, 2014(11): 123-129. (in Chinese) doi: 10.3969/j.issn.1002-0268.2014.11.020 [11] CHENG Y H, CHEN S Y. Perceived accessibility, mobility, and connectivity of public transportation systems[J]. Transportation Research PartA: Policy and Practice, 2015(77): 386-403. [12] CHENG Y H, LIU K C. Evaluating bicycle-transit users'perceptions of intermodal inconvenience[J]. Transportation Research Part A: Policy and Practice, 2012, 46(10): 1690-1706. doi: 10.1016/j.tra.2012.10.013 [13] SARRIAS M. Discrete choice models with random parameters in R: the GMNL package[J]. Journal of Statistical Software, 2016, 74(10): 1-31. [14] MCFADDEN D, TRAIN K. Mixed MNL models for discrete response[J]. Journal of Applied Econometrics, 2000, 15(5): 447-470. doi: 10.1002/1099-1255(200009/10)15:5<447::AID-JAE570>3.0.CO;2-1 [15] 韦杰. 考虑潜变量的电动汽车购买意愿选择模型研究[D]. 乌鲁木齐: 新疆大学, 2021.WEI J. Research on purchase intention choice model of electric vehicle considering latent variables[D]. Wulumuqi: Xinjiang University, 2021. (in Chinese) [16] 刘建荣, 郝小妮. 到达车站的步行时间对老年人公交选择的影响[J]. 交通运输系统工程与信息, 2020, 20(1): 124-129. https://www.cnki.com.cn/Article/CJFDTOTAL-YSXT202001021.htmLIU J R, HAO X N. Effect of bus stop walking time on elderly's bus choice[J]. Journal of Transportation Systems Engineering and Information Technology, 2020, 20(1): 124-129. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-YSXT202001021.htm [17] 刘志伟, 张荣堂, 刘建荣. 城市土地利用对居民通勤方式选择的影响[J]. 武汉理工大学学报(交通科学与工程版), 2020, 44(6): 989-993. doi: 10.3963/j.issn.2095-3844.2020.06.009LIU Z W, ZHANG R T, LIU J R. Influence of urban land use on resident's choice of commuting mode[J]. Journal of Wuhan University of Technology(Transportation Science & Engineering) 2020, 44 (6): 989-993. (in Chinese) doi: 10.3963/j.issn.2095-3844.2020.06.009 [18] DONG B, MA X, CHEN F, et al. Investigating the differences of single-vehicle and multivehicle accident probability using mixed logit model[J]. Journal of Advanced Transportation, 2018(2018): 2702360. [19] WANG W, YUAN Z, LIU Y, et al. A random parameter logit model of immediate red-light running behavior of pedestrians and cyclists at major-major intersections[J]. Journal of Advanced Transportation, 2019(2019): 2345903. [20] ZENG T. Essays on the random parameters logit model[D]. Baton Rouge: Louisiana State University, 2011. [21] 刘志伟, 刘建荣, 邓卫. 无人驾驶汽车对出行方式选择行为的影响[J]. 西南交通大学学报, 2021, 56(6): 1161-1168. https://www.cnki.com.cn/Article/CJFDTOTAL-XNJT202106004.htmLIU Z W, LIU J R, DENG W. Impact of autonomous vehicle on travel mode choice behavior[J]. Journal of Southwest Jiaotong University, 2021, 56(6): 1161-1168. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-XNJT202106004.htm [22] ECHANIZ E, DELL'OLIO L, IBEASÁ. Modelling perceived quality for urban public transport systems using weighted variables and random parameters[J]. Transport Policy, 2018(67): 31-39. [23] DAVIS F D. Perceived usefulness, perceived ease of use, and user acceptance of information technology[J]. MIS Quarterly, 1989, 38(3): 982-1003. [24] JI W, XIAO W. Structural equation modeling: Applications using mplus[M]. London: John Wiley & Son, 2012. [25] SARRIAS M, DAZIANO R. Multinomial logit models with continuous and discrete individual heterogeneity in R: The GMNL package[J]. Journal of Statistical Software, 2017, 79(2): 1-46. [26] 刘建荣, 郝小妮. 基于随机系数Logit模型的市内出行方式选择行为研究[J]. 交通运输系统工程与信息, 2019, 19(5): 108-113. https://www.cnki.com.cn/Article/CJFDTOTAL-YSXT201905015.htmLIU J R, HAO X N. Travel mode choice in city based on random parameters logit model[J]. Journal of Transportation Systems Engineering and Information Technology, 2019, 19(5): 108-113. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-YSXT201905015.htm [27] 刘志伟, 宋正沄, 邓卫, 等. 无人驾驶汽车对中短距离市际出行方式选择行为的影响[J]. 交通信息与安全, 2022, 40(2): 91-97. doi: 10.3963/j.jssn.1674-4861.2022.02.011LIU Z W, SONG Z Y, DENG W, et al. Impacts of autonomous vehicles on mode choice behavior in the context of short-and medium-distance intercity travel[J]. Journal of Transport Information and Safety, 2022, 40(2): 91-97. (in Chinese) doi: 10.3963/j.jssn.1674-4861.2022.02.011 -
点击查看大图
表(10)
计量
- 文章访问数: 907
- HTML全文浏览量: 301
- PDF下载量: 42
- 被引次数: 0