定制公交居民出行方式选择的影响因素

doi:10.3976/j.issn.1002-4026.2021.06.013

摘要/Abstract

摘要：

掌握影响定制公交选择概率的因素,对于定制公交的运营管理具有重要作用。结合广义Logit回归模型和条件Logit回归模型的混合模型,将居民个人属性与出行方案属性同时考虑进出行方式效用函数中。根据定制公交的在旅途距离及服务水平的设置定位,选取定制公交、出租车与常规公交作为出行选择集。运用问卷调查数据进行模型参数拟合,得到不同影响因素对于出行方式的边际效用和分布概率的影响。通过拟合结果,得到出行价格变量的系数在常规公交及定制公交1的效用函数中绝对值较大,分别为0.195 4和0.117 3,而在定制公交2和出租车的效用函数中绝对值较小,分别为0.156 1和0.057 1。行程时间变量在定制公交2和出租车效用函数中的绝对值较大,分别为0.073 5和0.080 7,在常规公交和定制公交1的效用函数中绝对值较小,分别为0.020 4和0.015 0。说明定制公交的服务需根据潜在乘客的出行价格及出行时间敏感度进行设置,研究结果对设置定制公交运营策略具有重要参考意义。

关键词: 出行方式选择, 定制公交, 效用函数, 混合模型, 边际效用

Abstract:

It is crucial for the operation and management of customized shuttle bus to acknowledge the factors influencing chosen rate. This study proposes a mixed model based on both generalized and conditional Logit models. Within the modeling structure, this study develops a utility function with the individual attributes of residents and travel scheme attributes. According to the travel distance and service level of customized shuttle bus, a set of customized shuttle bus, taxi, and public transit bus as options is determined. Based on stated preference data, the model estimation provides the marginal utility of and the distribution probability of the travel mode for different influencing factors. Results suggest that the absolute value of travel price variable coefficients in the utility function of the conventional bus and customized bus-1 are larger, which are 0.195 4 and 0.117 3, respectively,where in customized bus-2 and taxi is smaller, which are 0.156 1 and 0.057 1, respectively. The absolute values of travel time variables in the utility functions of the customized bus-2 and taxi are 0.073 5 and 0.080 7, respectively, where in conventional bus and customized bus-1 are 0.020 4 and 0.015 0, respectively. This implies that the customized bus service should be designed according to the travel price and travel time sensitivity of potential passengers. These results shed light on setting up customized bus operation strategies.

Key words: travel mode choice, customized shuttle bus, utility function, mixed model, marginal utility

中图分类号:

U491.1⁺22

张文波, 左玙璠, 王岳平, 陈景旭, 刘志远. 定制公交居民出行方式选择的影响因素[J]. 山东科学, 2021, 34(6): 97-105.

ZHANG Wen-bo, ZUO Yu-fan, WANG Yue-ping, CHEN Jing-xu, LIU Zhi-yuan. Factors influencing residents' travel mode selection considering customized shuttle bus[J]. Shandong Science, 2021, 34(6): 97-105.

图/表 3

表1

表2

表3

参考文献 27

[1]	梁安宁, 黄娜娜, 张兵, 等. 基于NL模型的昌九客运交通方式选择分析[J]. 华东交通大学学报, 2020, 37(1):54-60. DOI: 10.16749/j.cnki.jecjtu.2020.01.008. doi: 10.16749/j.cnki.jecjtu.2020.01.008
	LIANG A N, HUANG N N, ZHANG B, et al. Analysis on choice of Changjiu passenger transport mode based on NL model[J]. Journal of East China Jiaotong University, 2020, 37(1):54-60. DOI: 10.16749/j.cnki.jecjtu.2020.01.008. doi: 10.16749/j.cnki.jecjtu.2020.01.008
[2]	丰晓艳, 张亚伟. 基于交通出行行为的郑焦城际公交客流预测[J]. 山东科学, 2018, 31(1):79-87. DOI: 10.3976/j.issn.1002-4026.2018.01.014. doi: 10.3976/j.issn.1002-4026.2018.01.014
	FENG X Y, ZHANG Y W. Passenger flow forecast of Zhengzhou-Jiaozuo intercity bus based on traffic travel behavior[J]. Shandong Science, 2018, 31(1):79-87. DOI: 10.3976/j.issn.1002-4026.2018.01.014. doi: 10.3976/j.issn.1002-4026.2018.01.014
[3]	马姝宇, 邵春福. 基于NL模型的节假日绿色出行行为研究[J]. 山东科学, 2017, 30(1):76-81. DOI: 10.3976/j.issn.1002-4026.2017.01.012. doi: 10.3976/j.issn.1002-4026.2017.01.012
	MA S Y, SHAO C F. Study on holiday green travel behavior based on Nested Logit model[J]. Shandong Science, 2017, 30(1):76-81. DOI: 10.3976/j.issn.1002-4026.2017.01.012. doi: 10.3976/j.issn.1002-4026.2017.01.012
[4]	王旭, 吴建军. 基于累积前景理论的突发事件下轨道交通乘客路径选择研究[J]. 山东科学, 2015, 28(2):63-69. DOI: 10.3976/j.issn.1002-4026.2015.02.011. doi: 10.3976/j.issn.1002-4026.2015.02.011
	WANG X, WU J J. Cumulative prospect theory based urban railway traffic route choice for travelers in emergent events[J]. Shandong Science, 2015, 28(2):63-69. DOI: 10.3976/j.issn.1002-4026.2015.02.011. doi: 10.3976/j.issn.1002-4026.2015.02.011
[5]	张荣花, 关宏志, 赵磊, 等. 考虑汽车共享服务的出行方式选择影响因素研究[J]. 交通运输研究, 2020, 6(3):48-54. DOI: 10.16503/j.cnki.2095-9931.2020.03.006. doi: 10.16503/j.cnki.2095-9931.2020.03.006
	ZHANG R H, GUAN H Z, ZHAO L, et al. Influence factors of travel mode choice considering car sharing[J]. Transport Research, 2020, 6(3):48-54. DOI: 10.16503/j.cnki.2095-9931.2020.03.006. doi: 10.16503/j.cnki.2095-9931.2020.03.006
[6]	朱顺应, 陈秋成, 肖文彬, 等. 低碳补贴下的武汉市居民出行方式选择边际效应[J]. 安全与环境学报, 2020, 20(3):1118-1125. DOI: 10.13637/j.issn.1009-6094.2019.0910. doi: 10.13637/j.issn.1009-6094.2019.0910
	ZHU S Y, CHEN Q C, XIAO W B, et al. Marginal effect of Wuhan residents 'travel mode under low carbon subsidy scenario[J]. Journal of Safety and Environment, 2020, 20(3):1118-1125. DOI: 10.13637/j.issn.1009-6094.2019.0910. doi: 10.13637/j.issn.1009-6094.2019.0910
[7]	何晓平. 电力汽车共享条件下城市交通出行方式选择的演化博弈分析[D]. 兰州: 兰州交通大学, 2020.
	HE X P. Evolutionary game analysis of urban travel mode choices under electric vehicle sharing[D]. Lanzhou: Lanzhou Jiatong University, 2020.
[8]	李曼. 单车共享下城市居民出行方式选择行为模型分析:以湛江市为例[J]. 黑龙江交通科技, 2020, 43(8):212-214. DOI: 10.16402/j.cnki.issn1008-3383.2020.08.122. doi: 10.16402/j.cnki.issn1008-3383.2020.08.122
	LI M. Model analysis of urban residents' travel mode choice behavior under bicycle sharing: taking Zhanjiang City as an example[J]. Communications Science and Technology Heilongjiang, 2020, 43(8):212-214. DOI: 10.16402/j.cnki.issn1008-3383.2020.08.122. doi: 10.16402/j.cnki.issn1008-3383.2020.08.122
[9]	CECCATO R, CHICCO A, DIANA M. Evaluating car-sharing switching rates from traditional transport means through logit models and Random Forest classifiers[J]. Transportation Planning and Technology, 2021, 44(2):160-175. DOI: 10.1080/03081060.2020.1868084. doi: 10.1080/03081060.2020.1868084
[10]	PAPU CARRONE A, HOENING V M, JENSEN A F, et al. Understanding car sharing preferences and mode substitution patterns: A stated preference experiment[J]. Transport Policy, 2020, 98:139-147. DOI: 10.1016/j.tranpol.2020.03.010. doi: 10.1016/j.tranpol.2020.03.010
[11]	POLITIS I, FYROGENIS I, PAPADOPOULOS E, et al. Shifting to shared wheels: Factors affecting dockless bike-sharing choice for short and long trips[J]. Sustainability, 2020, 12(19):8205. DOI: 10.3390/su12198205. doi: 10.3390/su12198205
[12]	奴尔加马力·安尼瓦, 王令飞, 艾力·斯木吐拉. 基于Logistic模型的定制公交出行选择影响因素研究:以乌鲁木齐市为例[J]. 公路与汽运, 2019(6):23-25.
	NUERJIAMALI·ANNIWA, WANG L F, AILI·SIMUTULA. Research on influencing factors of custom bus trip choice based on logistic model: taking Urumqi City as an example[J]. Highways & Automotive Applications, 2019(6):23-25.
[13]	郭啸, 邓勇, 张婧. 空铁复合网络中旅客组合出行方式选择行为的模型研究[J]. 湖南师范大学自然科学学报, 2019, 42(6):77-85.
	GUO X, DENG Y, ZHANG J. Study on combined travel modes' choice model in a compound network of air and high-speed railway[J]. Journal of Natural Science of Hunan Normal University, 2019, 42(6):77-85.
[14]	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. DOI: 10.16097/j.cnki.1009-6744.2019.05.015. doi: 10.16097/j.cnki.1009-6744.2019.05.015
	LIU 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. DOI: 10.16097/j.cnki.1009-6744.2019.05.015. doi: 10.16097/j.cnki.1009-6744.2019.05.015
[15]	曾思萌, 叶茂, 刘英舜, 等. 共享单车引入后通勤者出行方式选择影响因素建模[J]. 交通信息与安全, 2019, 37(1):104-112.
	ZENG S M, YE M, LIU Y S, et al. A model of influencing factors of travel mode selection for commuters with bike-sharing[J]. Journal of Transport Information and Safety, 2019, 37(1):104-112.
[16]	WANG K, YE X, PENDYALA R M, et al. On the development of a semi-nonparametric generalized multinomial logit model for travel-related choices[J]. PLoS One, 2017, 12(10):e0186689. DOI: 10.1371/journal.pone.0186689. doi: 10.1371/journal.pone.0186689
[17]	王志勇, 叶祥松, 李飞星. 农业企业区位嵌入条件logit模型分析[J]. 统计与决策, 2018, 34(12):103-106. DOI: 10.13546/j.cnki.tjyjc.2018.12.025. doi: 10.13546/j.cnki.tjyjc.2018.12.025
	WANG Z Y, YE X S, LI F X. Analysis on Logit model of location embedding condition of agricultural enterprise[J]. Statistics & Decision, 2018, 34(12):103-106. DOI: 10.13546/j.cnki.tjyjc.2018.12.025. doi: 10.13546/j.cnki.tjyjc.2018.12.025
[18]	乐意. 基于混合选择模型的出行方式选择行为研究[C]// 中国城市规划学会城市交通规划学术委员会.品质交通与协同共治:2019年中国城市交通规划年会论文集.成都:中国城市规划设计研究院城市交通专业研究院, 2019: 14.
	LE Y. Research on travel mode selection behavior based on mixed selection model[C]// Academic committce of urban traffic planning of China urban planning society.Quality transportation and collaborative governance:Proceedings of 2019 China annual conference on urban traffic planning.Chengdu:Institute of urban traffic.China urban planning and design institute. 2019: 14.
[19]	GOKASAR I, GUNAY G. Modeling and implications of airport ground egress mode choice[J]. KSCE Journal of Civil Engineering, 2021, 25(2):669-679. DOI: 10.1007/s12205-020-5523-8. doi: 10.1007/s12205-020-5523-8
[20]	FERREIRA G E, HOWARD K, ZADRO J R, et al. People considering exercise to prevent low back pain recurrence prefer exercise programs that differ from programs known to be effective: a discrete choice experiment[J]. Journal of Physiotherapy, 2020, 66(4):249-255. DOI: 10.1016/j.jphys.2020.09.011. doi: 10.1016/j.jphys.2020.09.011
[21]	LEE B J. An introduction to mixed logit model[J]. Journal of Research Methodology, 2017, 2(1):79-104. DOI: 10.21487/jrm.2017.05.2.1.79. doi: 10.21487/jrm.2017.05.2.1.79
[22]	MA S H, YU Z L, LIU C Q. Nested logit joint model of travel mode and travel time choice for urban commuting trips in Xi'an, China[J]. Journal of Urban Planning and Development, 2020, 146(2):04020020. DOI: 10.1061/(asce)up.1943-5444.0000574. doi: 10.1061/(asce)up.1943-5444.0000574
[23]	范琪, 王炜, 杨洋, 等. 家庭收入差异对出行方式选择的影响分析[J]. 交通信息与安全, 2019, 37(6):111-120.
	FAN Q, WANG W, YANG Y, et al. Effects of household income differences on travel mode choice[J]. Journal of Transport Information and Safety, 2019, 37(6):111-120.
[24]	骆晨, 李想, 钟林峰, 等. 基于家庭属性差异的大学生出行方式选择行为研究[J]. 交通运输系统工程与信息, 2020, 20(3):227-232. DOI: 10.16097/j.cnki.1009-6744.2020.03.034. doi: 10.16097/j.cnki.1009-6744.2020.03.034
	LUO C, LI X, ZHONG L F, et al. Choice of travel mode for college students based on family attributes[J]. Journal of Transportation Systems Engineering and Information Technology, 2020, 20(3):227-232. DOI: 10.16097/j.cnki.1009-6744.2020.03.034. doi: 10.16097/j.cnki.1009-6744.2020.03.034
[25]	梁文博, 刘晨, 王艳玲. 基于Logistic回归模型的居民出行方式影响因素分析[J]. 生产力研究, 2020(1):125-127. DOI: 10.19374/j.cnki.14-1145/f.2020.01.030. doi: 10.19374/j.cnki.14-1145/f.2020.01.030
	LIANG W B, LIU C, WANG Y L. Analysis of factors influencing residents' travel modes based on Logistic regression model[J]. Productivity Research, 2020(1):125-127. DOI: 10.19374/j.cnki.14-1145/f.2020.01.030. doi: 10.19374/j.cnki.14-1145/f.2020.01.030
[26]	LI W B, KAMARGIANNI M. Steering short-term demand for car-sharing:a mode choice and policy impact analysis by trip distance[J]. Transportation, 2020, 47(5):2233-2265. DOI: 10.1007/s11116-019-10010-0. doi: 10.1007/s11116-019-10010-0
[27]	AZIMI G, RAHIMI A, ASGARI H, et al. Role of attitudes in transit and auto users' mode choice of ridesourcing[J]. Transportation Research Record: Journal of the Transportation Research Board, 2020, 2674(11):1-16. DOI: 10.1177/0361198120941841. doi: 10.1177/0361198120941841

变量类型	变量符号	变量含义	变量处理
个人基本属性	GEN	性别	GEN=1 男性
	AGE	年龄,分为5个等级,以18岁以下为参考	AGE_1=1 (18,25]
			AGE_2=1 (25,35]
			AGE_3=1 (35,50]
			AGE_4=1 (50, +∞)
	LINC	是否拥有驾驶证,以无驾驶证为参考	LINC=1 有驾驶证
	IND_INC	个人平均月收入,共分为5个等级,以收入(0,3000]为参考,单位为元	IND_INC_1=1 (3000,6000]
			IND_INC_2=1 (6000,10 000]
			IND_INC_3 =1 (10 000,15 000]
			IND_INC_4=1 (15 000, +∞)
	HOU_INC	家庭平均月收入,共5个等级,以收入为(0,5000]为参考,单位为元	HOU_INC_1=1 (5000,10 000]
			HOU_INC_2=1 (10 000,15 000]
			HOU_INC_3=1 (15 000,20 000]
			HOU_INC_4=1 (20 000, +∞)
	CAR_OWN	家庭私家车拥有量,以私家车拥有量为0为参考	CAR_OWN_1=1 拥有1辆私家车
	CAR_OWN	家庭私家车拥有量,以私家车拥有量为0为参考	CAR_OWN _2=1 拥有2辆及以上私家车
变量类型	变量符号	变量含义	变量处理
出行行为属性	PURP	出行目的,以通勤为参考	PURP=0代表出行目的为通勤
出行方案属性	SEAT	是否有座位,以座位不确定为参考	SEAT=1 确定有座位
	TRAN	是否需要换乘,分为3类,以不需要换乘为参考	TRAN_1=1 需换乘1次
	TRAN	是否需要换乘,分为3类,以不需要换乘为参考	TRAN_2=1 需换乘2次
	WAIT_T	上车前的等待时间,以2~5min为参考,单位为 min	WAIT_T_1=1 (5,10]
			WAIT_T_2=1 (10,15]
			WAIT_T_3=1 (15, +∞)
	WALK_D	至乘车点的步行距离,分为5个等级,以步行(0,500]为参考,单位为 m	WALK_D_1=1 步行(500,1000]
	WALK_D	至乘车点的步行距离,分为5个等级,以步行(0,500]为参考,单位为 m	WALK_D_2=1 步行(1000, +∞)
	BIAS	到达时间偏差,分为6个等级,以准时到达为参考	BIAS_1=1 早到2 min
			BIAS_2=1 早到5 min
			BIAS_3=1 早到6 min
			BIAS_4=1 早到8 min
			BIAS_5=1 早到10 min
	FEE	出行花费	连续变量
	TIME	行程时间	连续变量

出行方式	出行费用/元	出行时间/min
常规公交	-0.195 4	-0.020 4
出租车	-0.057 1	-0.080 7
定制公交1	-0.117 3	-0.015 0
定制公交2	-0.156 1	-0.073 5

影响因素	含义	取值
换乘	以不换乘为参考,换乘一次	-0.094 6
到达时间偏差 (以准时到达为参考)	早到2 min	0.167 4
	早到5 min	0.174 8
	早到6 min	0.200 2
	早到8 min	0.515 1
	早到10 min	1.135 4