Behavior Preference in the Decisions-making Process for Streetcar Development Considering Heterogeneity Within the Clusters of Urban Residents
-
摘要: 将有轨电车发展的研讨视为1项决策过程,按照基本属性、出行目的、出行模式等异质性特征,将出行群体划分为不同群组,开展基于不同异质性群组对发展有轨电车决策的偏好特性研究。通过融合行为偏好(RP)的意向偏好(SP)问卷调查,获取不同群组的基本属性及决策偏好特征数据。考虑有轨电车发展决策同时受环境要素、个体社会经济属性及出行需求特征等多层级因素的影响,对不同出行群组在不同情景下的决策偏好数据进行了多次测量,并引入多水平Logistic模型构建了考虑群组异质性的有轨电车发展决策偏好模型。选取了公共交通通勤出行、公共交通非通勤出行及非公共交通出行这3类异质性群组,对有轨电车发展决策偏好模型进行了参数估计。结果表明:①个体对有轨电车技术特性的感知并不会对有轨电车发展决策产生影响;②同一异质性群组,年龄的增长对有轨电车发展决策负向影响逐渐减小,说明年龄越大个体对有轨电车发展决策偏好逐渐增强,这种趋势在公共交通非通勤出行群组中更为显著;③可支配小汽车数对所有异质性群组的有轨电车发展决策偏好均呈显著负向影响,即家庭可支配小汽车数量越多,个体对有轨电车发展决策的支持度会越低;④出行时间及成本属性对有轨电车发展的决策偏好均有显著负向影响,即随着出行时间或出行成本的增加,对有轨电车发展决策的支持呈降低趋势,且该负向影响在骨架交通功能感知下较品质交通功能感知下更为敏感。
-
关键词:
- 交通信息 /
- 有轨电车 /
- 群组异质性 /
- 多水平Logistic模型 /
- 决策偏好
Abstract: Considering the survey for streetcar development as a part of decision-making process, travelers are divided into different clusters according to their heterogeneous characteristics such as personal attributes, travel purposes, and travel modes. A study on decisions-making preference to streetcar development based on the cluster heterogeneity is then carried out. Based on a questionnaire survey with Stated Preference (SP) and Revealed Preference (RP), data of personal attributes and characteristics of decision-making preference of different clusters are collected. The data has been obtained with multiple measurements in different travel groups and scenarios, considering the influences of multi-level factors, i.e. environmental factors, personal socio-economic attributes, and travel demands, on behavioral preference in the decisions-making process for streetcar development. A multilevel logistic model is used to develop a model with cluster heterogeneity. Parameters of the model are estimated using three heterogeneous cluster samples, i.e. commuting scenario with public transportation, traveling other than commuting scenario with public transportation, and traveling with private transportation scenario. The results show that: ①Individual's perception of tram technology have no impact on the preference for its development.②Negative effect of increasing age on the preference for tram development gradually decreases in the same heterogeneous group, which indicates the preference for tram development gradually increases with age, and this trend is more significant in the traveling other than commuting scenario of public transportation.③The number of available cars has significant negative influence on the preference of tram development in all heterogeneous groups. Study results show that the higher the number of available cars in the household, the lower the support of individuals for tram development. ④Both travel time and cost have significant negative effects on preferences for tram development, as the support for tram development tends to decrease as travel time or travel cost increases, and it is also found that travel groups are more sensitive such a negative impact under the perception of the skeletal transportation function than that under the perception of quality transportation function. -
表 1 问卷样本类型细分表
Table 1. Table of the sample types
个人属性 类型 通勤情景 非通勤情景 样本量 首选公共交通 次选公共交通 首选公共交通 次选公共交通 通勤者 I型 √ √ 162 Ⅱ型 √ √ 142 Ⅲ型 √ √ 56 Ⅳ型 √ √ 77 Ⅴ型 ✘ ✘ √ 48 Ⅵ型 ✘ ✘ √ 53 Ⅶ型 √ ✘ ✘ 248 Ⅷ型 √ ✘ ✘ 195 Ⅸ型 ✘ ✘ ✘ ✘ 392 非通勤者 Ⅹ型型 √ 73 Ⅺ型 √ 54 Ⅻ型 ✘ ✘ 38 表 2 考虑群组异质性的样本统计
Table 2. Sample statistics considering cluster heterogeneity
类型 公共交通通勤出行 公共交通非通勤出行 非公共交通出行 样本量 880 665 430 表 3 样本描述性统计分析
Table 3. Descriptive statistical analysis of the sample
样本变量 总样本比例/% 样本细分类别 χ2检验 公共交通通勤出行样本比例/% 公共交通非通勤出行样本比例/% 非公共交通出行样本比例/% 性别 男 51.3 50.4 50.7 53.6 p=0.016 女 48.7 49.6 49.3 46.4 年龄/岁 ≥ 18~25 18.8 19.6 19.8 16.3 p < 0.01 > 25~30 42.0 41.5 41.7 43.2 > 30~40 23.5 23.0 21.4 26.8 > 40~60 12.7 12.5 13.2 12.5 > 60 3.0 3.4 3.9 1.2 教育程度 高中及以下 8.3 8.5 8.4 8.0 p < 0.01 大专或本科 64.2 64.8 63.5 63.8 硕士及以上 27.5 26.7 28.1 28.2 所在城市 县级市 6.2 6.2 6.5 5.9 p=0.146 直辖市 26.5 25.5 26.7 28.3 省会城市 59.5 60.9 58.2 58.3 地级市 7.8 7.4 8.6 7.5 城市人口规模 ≤ 300万 7.1 7.3 7.1 6.6 p=0.306 > 300万~500万 11.6 11.5 12.3 10.8 > 500万~1 000万 49.7 47.8 48.6 54.4 > 1 000万 31.7 33.4 32.0 28.2 职业性质 政府机关 26.1 27.6 18.4 32.1 p=0.017 规划设计院所 22.9 22.2 23.4 23.7 其他企业公司 16.2 15.8 14.6 18.8 个体经营 8.3 6.1 7.3 13.5 学生 21.1 28.3 20.1 8.7 离退休或无业 5.4 0.0 16.2 3.2 平均收人/元 ≤ 3 000 17.4 19.5 19.2 11.4 p < 0.01 > 3000~6 000 20.0 22.5 21.7 13.3 > 6000~8 000 27.7 25.2 25.6 34.8 > 8000~10 000 19.5 17.2 18.6 24.8 > 10 000 15.4 15.6 14.9 15.7 家庭组成/人 1 12.3 13.1 12.6 10.5 p=0.069 2 16.9 19.4 16.8 12.3 3 37.3 36.2 35.5 41.6 ≥ 4 33.5 31.3 35.1 35.6 表 4 考虑群组异质性的模型变量定义及赋值
Table 4. Definition and assignment of variables in the preference model
变量 定义 赋值 结果变量 y 是否支持发展有轨电车 是=1,否=0 水平层1解释变量 x1 性别 男=1,女=2 x2 年龄/岁 ≤ 25=1,> 25~30=2, > 30~ 40=3,> 40~60=4,> 60=5 x3 学历 高中及以下=1,大专或本科=2,硕士及以上=3 x4 所在城市类型 县级市=1,地级市=2, 省会城市=3, 直辖市=4 x5 所在城市人口规模/万人 ≤ 300=1,> 300~500=2,> 500~1 000=3, > 1 000=4 x6 平均月收人/元 ≤ 3 000=1,> 3 000~6 000= 2,> 6 000~8 000=3,> 8 000~ 10 000=4, > 10 000=5 x7 出行灵活度 完全不自由=1,有一定自由度= 2, 完全自由=3 x8 家庭成员人数 1=1,2=2, 3=3, ≥ 4=4 x9 可支配小汽车数 0=1,1=2, ≥ 2=3 x10 可支配电动车数 0=1,1=2, ≥ 2=3 x11 可支配自行车数 0=1,1=2, ≥ 2=3 x12 公交车周乘坐频率 0=1, 1~5=2, 6~10=3, >10=4 x13 共享单车周使用频率 0=1, 1~5=2, 6~10=3, >10=4 x14 出租车周使用频率 0=1, 1~5=2, 6~10=3, >10=4 x15 出行时间 数值 x16 出行成本 数值 水平层2解释变量 u1 对功能定位的感知 骨架交通感知=-1,辅助交通感知=0, 品质交通感知=1 水平层3解释变量 g1 对技术特性的感知 建设特性感知=1, 运营特性感知=2 表 5 有轨电车发展决策偏好空模型参数估计结果
Table 5. Estimation results of decision preference null model parameters
参数 异质性群组1 异质性群组2 异质性群组3 估计值 P 估计值 P 估计值 P 固定部分 截距 1.324 0.000 0.819 0.000 0.572 0.000 随机部分 水平层3方差 0.023 0.618 0.042 1.401 0.036 1.216 水平层2方差 0.906 0.000 0.767 0.000 0.505 0.000 水平层3 ICC 0.011 0.004 0.005 水平层2 ICC 0.176 0.182 0.136 表 6 完整模型估计结果
Table 6. Results of model estimation
参数 异质性群组1 异质性群组2 异质性群组3 估计值 标准误差 P 估计值 标准误差 P 估计值 标准误差 P 固定效应部分 水平层1解释变量 截距 1.646 0.122 0.000 0.331 0.106 0.000 2.502 1.312 0.000 x1=2 0.180 0.077 0.000 x2=2 -0.221 0.119 0.027 -0.152 0.136 0.011 x2=3 -0.409 0.123 0.014 -0.246 0.122 0.023 x2=4 -0.277 0.092 0.021 -0.135 0.088 0.005 x2=5 -0.202 0.153 0.018 -0.128 0.113 0.008 x4=2 0.225 0.121 0.015 0.269 0.064 0.016 0.097 0.163 0.015 x4=3 0.174 0.118 0.003 0.206 0.081 0.012 0.085 0.158 0.011 x4 =4 0.092 0.160 0.008 0.155 0.025 0.019 0.044 0.206 0.018 x5=2 0.231 0.098 0.015 0.327 0.073 0.017 0.173 0.105 0.003 x5=3 0.166 0.124 0.016 0.202 0.086 0.031 0.144 0.093 0.000 x5=4 0.071 0.165 0.035 0.176 0.080 0.022 0.108 0.114 0.009 x6=2 -0.097 0.119 0.355 -0.026 0.078 0.169 x6=3 -0.124 0.106 0.196 -0.071 0.077 0.351 x6=4 -0.156 0.219 0.029 -0.176 0.065 0.003 x6=5 -0.281 0.108 0.026 -0.184 0.094 0.009 x7=2 0.139 0.062 0.009 (x7=2) · u1 0.102 0.046 0.003 x7 =3 0.178 0.059 0.037 x9=2 -0.241 0.139 0.014 -0.258 0.114 0.000 -0.552 0.163 0.001 (x9=2) · u1 0.188 0.127 0.007 0.165 0.129 0.000 0.313 0.111 0.000 x9=3 -0.260 0.171 0.004 -0.276 0.175 0.001 -0.615 0.193 0.007 x12=2 0.128 0.116 0.000 0.081 0.147 0.000 x12=3 0.196 0.082 0.001 0.363 0.125 0.000 x12=4 0.272 0.122 0.002 0.435 0.161 0.000 (x12=4) · u1 0.211 0.124 0.000 x15 -3.633 0.127 0.000 -3.486 0.199 0.000 -4.726 0.129 0.000 x15 · u1 0.761 0.091 0.000 1.725 0.172 0.000 0.882 0.136 0.000 x16 -2.879 0.139 0.000 -3.098 0.171 0.000 -2.611 0.188 0.000 x15 · u1 0.842 0.155 0.000 1.449 0.168 0.000 0.725 0.164 0.000 水平层2解释变量 u1 0.225 0.104 0.000 0.392 0.149 0.000 0.108 0.011 0.000 随机效应部分 截距, σε0j2 0.330 0.182 0.000 0.292 0.083 0.000 0.245 0.048 0.031 x7=2,σε1j2 0.408 0.226 0.001 x9=2,σε2j2 0.411 0.162 0.000 0.608 0.209 0.001 0.352 0.183 0.000 x12=4,σε3j2 0.283 0.163 0.000 x15,σε4j2 0.695 0.242 0.000 0.361 0.152 0.001 0.415 0.144 0.000 x16σε5j2 0.525 0.195 0.000 0.408 0.166 0.000 0.218 0.138 0.000 表 7 有轨电车发展决策相对偏好度RPI变化量
Table 7. RPI change of relative preference degree
参数 异质性群组1 异质性群组2 异质性群组3 x1=2 0.197* x2=2 -0.198* -0.141* x2=3 -0.336* -0.218* x2=4 -0.242* -0.126* x2=5 -0.183* -0.120* x4=2 0.252* 0.309* 0.102* x4=3 0.190* 0.229* 0.089* x4=4 0.096* 0.168* 0.045* x5=2 0.260* 0.387* 0.189* x5=3 0.181* 0.224* 0.155* x5=4 0.074* 0.192* 0.114* x6=2 -0.092 -0.026 x6=3 -0.117 -0.069 x6=4 -0.144* -0.161* x6=5 -0.245* -0.168* x7=2 0.149* (x7=2) · u1 0.107* x7=3 0.195* x9=2 -0.214* -0.227* -0.424* (x9=2) · u1 0.207* 0.179* 0.368* x9=3 -0.229* -0.241* -0.459 x12=2 0.137* 0.084* x12=3 0.217* 0.438* x12=4 0.313* 0.545* (x12=4) · u1 0.235* x15 -0.974* -0.969* -0.991* x15 · u1 1.140* 4.613* 1.416* x16 -0.944* -0.955* -0.927* x15 · u1 1.321* 3.259* 1.065* u1 0.252* 0.480* 0.114* 注:*表示p<0.05,说明差异有统计学意义。 表 8 出行时间与出行成本对决策偏好度影响的单位变化值
Table 8. Unit change values of the effect of travel time and travel cost on decision preference degree
参数 不同功能定位感知情景下 模式1 模式2 模式3 出行时间 异质性群组1 -0.987 6 -0.973 6 -0.943 4 异质性群组2 -0.994 5 -0.969 4 -0.828 1 异质性群组3 -0.996 3 -0.991 1 -0.978 6 出行成本 异质性群组1 -0.975 8 -0.943 8 -0.869 6 异质性群组2 -0.989 4 -0.9549 -0.807 8 异质性群组3 -0.964 4 -0.926 5 -0.848 3 -
[1] 马莹莹, 陆思园, 张晓明, 等. 考虑个体风险偏好差异的高速公路出行选择模型[J]. 吉林大学学报(工学版), 2021, 51 (5): 1673-1683. https://www.cnki.com.cn/Article/CJFDTOTAL-JLGY202105014.htmMA Y Y, LU S Y, ZHANG X M, et al. Model of highway travel selection considering individual risk preference difference[J]. Journal of Jilin University(Engineering and Technology Edition), 2021, 51(5): 1673-1683.( in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-JLGY202105014.htm [2] 李军, 贾顺平, 王瑜琼, 等. 旅客购票模式与渠道关联选择行为的建模分析[J]. 交通运输系统工程与信息, 2019, 19(3): 176-181+194. https://www.cnki.com.cn/Article/CJFDTOTAL-YSXT201903026.htmLI J, JIA S P, WANG Y Q, et al. Modeling analysis of related choice for passenger ticketing mode and channel[J]. Journal of Transportation Systems Engineering and Information Technology, 2019, 19(3): 176-181+194.( in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-YSXT201903026.htm [3] 姚恩建, 陈伟迪, 卢天伟, 等. 考虑出行者选择偏好的出行方式选择模型[J]. 北京交通大学学报, 2020, 44(1): 42-48. doi: 10.11860/j.issn.1673-0291.20190072YAO E J, CHEN W D, LU T W, et al. Transportation mode selection model considering traveler's personal preferences[J]. Journal of Beijing Jiaotong University, 2020, 44(1): 42-48.( in Chinese) doi: 10.11860/j.issn.1673-0291.20190072 [4] 刘少博, 张鑫泽, 李艺玮, 等. 地铁站人员疏散路径选择行为分析及Logit建模[J]. 交通信息与安全, 2019, 37(2): 91-98. doi: 10.3963/j.issn.1674-4861.2019.02.013LIU S B, ZHANG X Z, LI Y W, et al. Analyses and logit models of route choice behaviors during metro station evacuation[J]. Journal of Transport Information and Safety, 2019, 37(2): 91-98.( in Chinese) doi: 10.3963/j.issn.1674-4861.2019.02.013 [5] 左利兴. 多层LOGIT模型的理论分析研究[J]. 道路交通与安全, 2010, 10(2): 27-30. https://www.cnki.com.cn/Article/CJFDTOTAL-DLJA201002008.htmZUO L X. Theoretical analysis of Nest Logit Model[J]. Journal of Transportation Engineering, 2010, 10(2): 27-30. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-DLJA201002008.htm [6] 胡严艺, 蒲政, 王沛. 基于Logit模型的碳排放收费对居民出行方式选择的研究[J]. 交通运输工程与信息学报, 2018, 16(4): 57-62. doi: 10.3969/j.issn.1672-4747.2018.04.008HU Y Y, PU Z, WANG P. Study on the impacts of traffic carbon emission pricing on resident trip behavior using logit model[J]. Journal of Transportation Engineering and Information, 2018, 16(4): 57-62.( in Chinese) doi: 10.3969/j.issn.1672-4747.2018.04.008 [7] 张新洁, 关宏志, 赵磊, 等. 有限理性视野下出行者出行方式选择分层Logit模型研究[J]. 交通运输系统工程与信息, 2018, 18(6): 110-116. https://www.cnki.com.cn/Article/CJFDTOTAL-YSXT201806016.htmZHANG X J, GUAN H Z, ZHAO L, et al. Nested logit model on travel mode choice under bounded rational view[J]. Journal of Transportation Systems Engineering and Information Technology, 2018, 18(6): 110-116.( in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-YSXT201806016.htm [8] 李亚平. 价格调控策略对城市居民出行方式决策行为的影响研究[D]. 南京: 东南大学, 2018.LI Y P. Impacts of pricing strategies on travel mode choice behavior of urban resident[D]. Nanjing: Southeast University, 2018.( in Chinese) [9] 姚丽亚, 孙立山, 关宏志. 基于分层Logit模型的交通方式选择行为研究[J]. 武汉理工大学学报(交通科学与工程版), 2010, 34(4): 738-741. doi: 10.3963/j.issn.1006-2823.2010.04.023YAO L Y, SUN L S, GUAN H Z. Study on modal split method based on nested logit model[J]. Journal of Wuhan University of Technology(Transportation Science & Engineering), 2010, 34(4): 738-741.( in Chinese) doi: 10.3963/j.issn.1006-2823.2010.04.023 [10] 程亚鹏. 住房市场嵌套结构与价格变动的多层模型实证研究[J]. 系统工程, 2018, 36(8): 71-78. doi: 10.3969/j.issn.1001-2362.2018.08.046CHENG Y P. Empirical research on the nested housing market structure and price variation based on a multilevel model[J]. Systems Engineering, 2018, 36(8): 71-78.( in Chinese) doi: 10.3969/j.issn.1001-2362.2018.08.046 [11] 黄晓燕, 曹小曙, 殷江滨, 等. 城市轨道交通和建成环境对居民步行行为的影响[J]. 地理学报, 2020, 75(6): 1256-1271. https://www.cnki.com.cn/Article/CJFDTOTAL-DLXB202006013.htmHUANG X Y, CAO X S, YIN J B, et al. The influence of urban transit and built environment on walking[J]. Acta Geographica Sinica, 2020, 75(6): 1256-1271.( in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-DLXB202006013.htm [12] SOHN S Y, CHOI H. Random effects logistic regression model for data envelopment analysis with correlated decision making units[J]. Journal of the Operational Research Society, 2017, 57(5): 552-560. [13] EUN J, JIYEONG K, HYUNJUNG K. Does environmental walkability matter? The role of walkable environment in active commuting[J]. International Journal of Environmental Research and Public Health, 2020, 17(4): 1261 doi: 10.3390/ijerph17041261 [14] ZHAO P, LI S. Bicycle-metro integration in a growing city: The determinants of cycling as a transfer mode in metro station areas in Beijing[J]. Transportation Research Part A: Policy & Practice, 2017(99): 46-60. [15] 吴密霞. 混合效应模型估计理论及方法[D]. 北京: 北京工业大学, 2004.WU M X. Mixed effect model estimation theory and method[D]. Beijing: Beijing University of Technology, 2004. (in Chinese) [16] 王济川, 谢海义, 姜宝法. 多层统计分析模型: 方法与应用[M]. 北京: 高等教育出版社, 2008.WANG J C, XIE H Y, JIANG B F. Multilevel models: methods and applications[M]. Beijing: Higher Education Press, 2008.( in Chinese) [17] CZADO C, PROKOPENKO S. Modelling transport mode decisions using hierarchical logistic regression models with spatial and cluster effects[J]. Statistical Modelling an Interna tional Journal, 2008, 8(4): 315-345. doi: 10.1177/1471082X0800800401 [18] GREEN P J. Penalized likelihood for general semi-paramet-ric regression models[J]. International Statistical Review, 1987, 55, 245-259. doi: 10.2307/1403404 [19] ROBINSON G K. That BLUP is a good thing: the estimation of random effects[J]. Statistical Science, 1991(6): 15-51. [20] RAMOS L E, BROWN J, NIXON H. The streetcar resurgence in the United States: Transit strategy, growth-machine tactic, or some of both?[J]. Transportation Research Record Journal of the Transportation Research Board, 2016(2540): 30-38. [21] RAMOS L.E., BROWN J, NIXON H. The Transit Performance of Modern-Era Streetcars: A Consideration of Five U. S. Cities[J]. Transportation Research Record Journal of the Transportation Research Board, 2015(2534): 57-67. [22] 孙嘉. 基于功能分析的现代有轨电车适应性后评估研究[D]. 南京: 东南大学, 2019.SUN J. Post-evaluation of adaptability of modern trams based on functional analysis[D]. Nanjing: Southeast University, 2019.( in Chinese) [23] 邴雪, 张玉一, 王英杰, 赵一北. 基于功能定位的有轨电车运营提升策略研究[J]. 交通工程, 2020, 20(6): 69-74.BING X, ZHANG Y Y, WANG Y J, et al. Study on the strategy of tram operation improvement based on functional positioning[J]. Journal of Transportation Engineering, 2020, 20(6): 69-74.( in Chinese) [24] 黎冬平. 有轨电车在多层次轨道交通中的应用思考[J]. 城市轨道交通, 2021(6): 43-45. https://www.cnki.com.cn/Article/CJFDTOTAL-CSGD202106021.htmLI D P. Reflections on the application of trams in multi-level rail transportation[J]. Urban Mass Transit, 2021(6): 43-45. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-CSGD202106021.htm
计量
- 文章访问数: 1163
- HTML全文浏览量: 456
- PDF下载量: 37
- 被引次数: 0