吉林大学学报(工学版) ›› 2021, Vol. 51 ›› Issue (4): 1261-1268.doi: 10.13229/j.cnki.jdxbgxb20200390

• 交通运输工程·土木工程 • 上一篇    

基于潜在类别的无人驾驶汽车选择行为

刘志伟1(),刘建荣2(),邓卫3   

  1. 1.武汉轻工大学 土木工程与建筑学院,武汉 430023
    2.华南理工大学 土木与交通学院,广州 510640
    3.东南大学 交通学院,南京 210096
  • 收稿日期:2020-06-03 出版日期:2021-07-01 发布日期:2021-07-14
  • 通讯作者: 刘建荣 E-mail:tonyliuzhiwei@whpu.edu.cn;ctjrliu@scut.edu.cn
  • 作者简介:刘志伟(1987-),男,讲师,博士. 研究方向:出行行为选择,公交服务质量. E-mail: tonyliuzhiwei@whpu.edu.cn
  • 基金资助:
    国家自然科学基金项目(51578247);湖北省自然科学基金项目(2020CFB290)

Travelers′ choice behavior of autonomous vehicles based on latent class

Zhi-wei LIU1(),Jian-rong LIU2(),Wei DENG3   

  1. 1.School of Civil Engineering and Architecture,Wuhan Polytechnic University,Wuhan 430023,China
    2.School of Civil Engineering and Transportation,South China University of Technology,Guangzhou 510640,China
    3.School of Transportation,Southeast University,Nanjing 210096,China
  • Received:2020-06-03 Online:2021-07-01 Published:2021-07-14
  • Contact: Jian-rong LIU E-mail:tonyliuzhiwei@whpu.edu.cn;ctjrliu@scut.edu.cn

摘要:

为了分析出行者出行选择偏好的异质性和无人驾驶汽车对出行选择行为的影响,基于扩展技术接受模型的理论框架,将心理潜变量纳入潜在类别条件Logit模型,建立混合选择模型进行实证研究。结果表明:相比传统的条件Logit模型,潜在类别条件Logit模型的拟合优度更高。受访出行者可以分为共享无人驾驶汽车偏好群体、小汽车偏好群体和私人无人驾驶汽车偏好群体,3个潜在类别分别占比为46.5%、13.0%和40.5%。小汽车偏好群体对步行与等待时间和出行费用持正向评价,私人无人驾驶汽车偏好群体对停车费用持负面评价。感知易用性和感知信任两个潜变量对出行者潜在类别划分具有显著影响。

关键词: 交通工程, 出行方式选择, 潜在类别条件Logit模型, 潜变量, 无人驾驶汽车

Abstract:

In order to analyze the preferences heterogeneity of travelers and the influence of autonomous vehicles on travel behavior, this study incorporates the latent psychological variables into the latent-class conditional Logit model based on the theoretical framework of the extended technology acceptance model. A hybrid choice model is established to conduct empirical research. The results show that compared with the traditional conditional Logit model, the latent-class conditional Logit model has higher fitness. Respondents can be divided into three subgroups: the shared autonomous vehicles preference subgroup, car preference subgroup, and private autonomous vehicles preference subgroup. The three subgroups account for 46.5%, 13.0% and 40.5%, respectively. Travelers from the car preference subgroup evaluate the walking and waiting time and travel costs positively, while travelers from the private autonomous vehicle preference subgroup evaluate parking costs negatively. The two latent variables of perceived ease of use and perceived trust have a significant influence on the classification of latent class.

Key words: traffic engineering, travel mode choice, latent-class conditional Logit model, latent variable, autonomous vehicle

中图分类号: 

  • U121

表1

表征心理潜在变量的显示变量"

潜在变量显示变量符号
感知有用(PU)无人驾驶车辆可以降低交通事故pu1
无人驾驶车辆可以降低交通拥堵pu 2
无人驾驶车辆可以降低交通成本pu 3
无人驾驶车辆可以改善驾驶行为表现pu 4
感知易用(PEU)学习使用无人驾驶车辆对我来说很简单peu 1
无人驾驶车辆的使用很清晰和容易理解peu 2
无人驾驶车辆的使用不需要太多的心理负担peu 3
感知信任(PT)无人驾驶车辆是可靠的pt1
无人驾驶车辆是值得信赖的pt 2
总之,我完全相信无人驾驶车辆pt 3
社会影响(SN)对我很重要的人支持我使用无人驾驶车辆sn1
对我很重要的人希望我将来能使用无人驾驶车辆sn 2
如果身边的人使用无人驾驶车辆,我也会使用的sn 3
行为意向(BIU)未来我会使用无人驾驶车辆biu1
未来我会购买无人驾驶车辆biu2
我会向亲朋好友推荐使用无人驾驶车辆biu3

图 1

假定情景示例"

表2

样本描述性统计"

变量表示符号定义百分比/%
性别GEND45.73
54.27
年龄AGE130岁及以下66.00
AGE231~45岁13.90
AGE346~55岁12.23
AGE455岁以上7.87
受教育程度EDU高中及以下10.05
大专11.39
本科66.67
硕士及以上11.89

家庭月收入

(元/月)

HINC15000以下24.29
HINC25000~1000041.71
HINC310000~2000021.60
HINC420000以上12.40

是否有学龄儿

CHILD53.10
46.90
是否拥有驾照LICENSE42.38
57.62

家庭是否拥有

小汽车

CAR63.99
36.01

是否拥有公交

IC卡

ICARD77.72
22.28
家庭总人口数HSIZE15.36
29.21
340.20
420.60
5人及以上24.62
通勤方式CAR2WORK使用小汽车上班30.65
BUS2WORK使用公交车上班18.09
TRAIN2WORK使用地铁上班21.94
OTHERS其他29.32

表3

样本数据的信度及效度检验"

变量题项KMO

因子

载荷

Cronbach′s Alpha
感知有用性(PU)pu 10.85070.9180.931
pu 20.927
pu 30.923
pu 40.876
感知易用性(EU)peu 10.77500.9560.956
peu 20.964
peu 30.954
感知信任(PT)pt 10.76800.9740.965
pt 20.972
pt 30.955
社会影响(SN)sn 10.76800.9740.965
sn 20.972
sn 30.955
行为意向(BIU)biu 10.77600.9790.932
biu 20.971
biu 30.968

表4

模型检验指标结果"

RMSEACFITLISRMR
参考值≤0.08≥0.900≥0.900<0.08
模型参数0.080.9760.9670.020

表5

模型CAIC及BIC比较"

类别CAICBIC
236823655
327312683
427842715
528482758
629282817

表6

潜在类别条件Logit模型回归系数和条件Logit模型回归系数"

ClassesVariableCoef.SEzP>z
choice1wwt-0.0950.029-3.320.001
ivt-0.0520.014-3.660
tc-0.0430.012-3.590
pc0.0330.0410.80.422
sasc2.2220.3057.30
casc-0.7010.326-2.150.031
choice2wwt0.1990.1001.990.047
ivt0.0200.0600.330.74
tc0.1520.0682.240.025
pc-0.0180.142-0.120.902
sasc2.7771.5291.820.069
casc4.3791.4443.030.002

choice3

wwt0.0900.0571.590.112
ivt-0.0120.020-0.60.551
tc-0.0030.021-0.170.867
pc-0.2100.082-2.540.011
sasc-4.9840.707-7.050
casc-2.8210.590-4.780
传统条件Logit模型wwt-0.0490.013-3.870
ivt-0.0160.005-3.080.002
tc-0.0080.005-2.020.043
pc-0.0600.019-3.240.001

表7

出行者特征对类别的影响"

ClassesVariableCoef.SEzP>z
share1GEND0.1270.2330.550.584
AGE11.4210.4273.330.001
OCCU-0.2690.131-2.060.039
HINC11.1490.2744.200
HINC2-0.6100.409-1.490.136
LICENSE-0.8550.259-3.300.001
CAR-0.1420.239-0.590.553
CAR2WORK-1.3490.294-4.590
TRAIN2WORK0.9700.2913.330.001
PU1.1930.8331.430.152
PEU-0.8930.339-2.630.008
SN0.2570.3070.840.403
PT0.7850.6301.250.213
BIU-1.5750.985-1.600.110
_cons1.2300.5262.340.019
share2GEND0.4520.4291.050.293
AGE11.9810.6812.910.004
OCCU-0.6420.220-2.910.004
HINC11.7460.4903.560
HINC23.0960.6554.720
LICENSE0.1560.4500.350.729
CAR0.0890.4720.190.85
CAR2WORK-1.1820.570-2.070.038
TRAIN2WORK-0.8910.597-1.490.135
PU-1.1841.223-0.970.333
PEU-0.7320.527-1.390.165
SN-0.0180.489-0.040.971
PT-0.7921.024-3.180.010
BIU-1.4161.493-0.950.343
_cons-2.1230.989-2.150.032

图2

年龄、月收入、是否开车上下班对潜在类别的影响"

1 Molin E, Mokhtarian P, Kroesen M. Multimodal travel groups and attitudes: a latent class cluster analysis of Dutch travelers[J]. Transportation Research Part A: Policy and Practice, 2016, 83(1): 14-29.
2 Zhang X, Guan H, Zhu H, et al. Analysis of travel mode choice behavior considering the indifference threshold[J]. Sustainability, 2019, 11(19): 1-23.
3 Boxall P C, Adamowicz W L. Understanding heterogeneous preferences in random utility models: a latent class approach[J]. Environmental and resource economics,2002, 23(4): 421-446.
4 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.
5 Ardeshiri A, Vij A. Lifestyles, residential location, and transport mode use: a hierarchical latent class choice model[J]. Transportation Research Part A: Policy and Practice,2019, 126(10): 342-359.
6 刘建荣,刘志伟. 基于出行者潜在类别的公交出行行为研究[J]. 华南理工大学学报:自然科学版, 2019, 47(6):119-126.
Liu Jian-rong, Liu Zhi-wei. Analysis of bus users' behavior based on latent class[J]. Journal of South China University of Technology (Natural Science Edition), 2019, 47(6): 119-126.
7 Ben-Akiva M, Mcfadden D, Gärling T, et al. Extended framework for modeling choice behavior[J]. Marketing Letters, 1999, 10(3): 187-203.
8 景鹏,隽志才,查奇芬. 考虑心理潜变量的出行方式选择行为模型[J]. 中国公路学报, 2014, 27(11): 84-92, 108.
Jing Peng, Zhi-cai Juan, Zha Qi-fen. Incorporating psychological latent variables into travel mode choice model[J]. China Journal of Highway Transportation, 2014, 27(11): 84-92, 108.
9 Davis F D, Bagozzi R P, Warshaw P R. User acceptance of computer technology: a comparison of two theoretical models[J]. Management science, 1989, 35(8): 982-1003.
10 Venkatesh V, Davis F D. A model of the antecedents of perceived ease of use: development and test[J]. Decision Sciences, 1996, 27(3): 451-481.
11 Mou J, Shin D, Cohen J F. Trust and risk in consumer acceptance of e-services[J]. Electronic Commerce Research, 2016,16(1):141.
12 Bansal P, Kockelman K M, Singh A. Assessing public opinions of and interest in new vehicle technologies: an Austin perspective[J]. Transportation Research Part C: Emerging Technologies, 2016, 67(6): 1-14.
13 Train K E. EM algorithms for nonparametric estimation of mixing distributions[J]. Journal of Choice Modelling, 2008, 1(1): 40-69.
14 Pacifico D, Yoo H I. Lclogit: a Stata command for fitting latent-class conditional Logit models via the expectation-maximization algorithm[J]. The Stata Journal, 2013, 13(3): 625-639.
15 Panagiotopoulos I, Dimitrakopoulos G. An empirical investigation on consumers' intentions towards autonomous driving[J]. Transportation Research Part C: Emerging Technologies, 2018, 95(10): 773-784.
16 Yáñez M F, Raveau S, Ortúzar J D D. Inclusion of latent variables in mixed Logit models: modelling and forecasting[J]. Transportation Research Part A: Policy and Practice,2010, 44(9): 744-753.
17 Wang J, Wang X. Structural Equation Modeling: Applications Using Mplus[M]. Chichester: John Wiley & Sons, 2012.
[1] 查伟雄,蔡其燕,李剑,严利鑫. 边路车辆出入条件下城市干线信号协调相位差优化[J]. 吉林大学学报(工学版), 2021, 51(2): 565-574.
[2] 栾鑫, 邓卫, 程琳, 陈新元. 特大城市居民出行方式选择行为的混合Logit模型[J]. 吉林大学学报(工学版), 2018, 48(4): 1029-1036.
[3] 孙轶轩, 邵春福, 岳昊, 朱亮. 基于SVM灵敏度的城市交通事故严重程度影响因素分析[J]. 吉林大学学报(工学版), 2014, 44(5): 1315-1320.
[4] 董宏辉,孙晓亮,贾利民,秦勇. 多模态的交通流量预测模型[J]. 吉林大学学报(工学版), 2011, 41(03): 645-649.
[5] 姜桂艳,吴超腾. 基于GPS采集车辆行程时间的路段划分模型[J]. 吉林大学学报(工学版), 2009, 39(增刊2): 177-0181.
[6] 吕能超,严新平,RanBin,徐堃. 基于道路出行者行为的大型活动停车模型[J]. 吉林大学学报(工学版), 2009, 39(增刊2): 135-0139.
[7] 宋现敏,王殿海,金盛. 感应式协调控制下绿信比优化方法[J]. 吉林大学学报(工学版), 2009, 39(增刊2): 154-0157.
[8] 王殿海,孙锋,金盛 . 两相位交叉口左转车通行能力计算方法[J]. 吉林大学学报(工学版), 2007, 37(04): 767-771.
[9] 李洪萍,裴玉龙,杨中良 .

快速路自由流速度及其影响因素

[J]. 吉林大学学报(工学版), 2007, 37(04): 772-776.
[10] 姜桂艳,郑祖舵,白竹,代磊磊,赵佳琪. 拥挤条件下可变信息板交通诱导信息对驾驶行为的影响[J]. 吉林大学学报(工学版), 2006, 36(02): 183-0187.
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!