吉林大学学报(工学版) ›› 2021, Vol. 51 ›› Issue (5): 1673-1683.doi: 10.13229/j.cnki.jdxbgxb20200409

• 交通运输工程·土木工程 • 上一篇    

考虑个体风险偏好差异的高速公路出行选择模型

马莹莹1,2(),陆思园2(),张晓明2,魏文术2   

  1. 1.华南理工大学 土木与交通学院,广州 510640
    2.广州市城市规划勘测设计研究院 交通规划设计所,广州 510060
  • 收稿日期:2020-06-11 出版日期:2021-09-01 发布日期:2021-09-16
  • 通讯作者: 陆思园 E-mail:mayingying@scut.edu.cn;1055054928@qq.com
  • 作者简介:马莹莹(1983-),女,副教授,博士.研究方向:城市交通系统分析.E-mail:mayingying@scut.edu.cn
  • 基金资助:
    广东省自然科学基金项目(2018A030313250)

Model of highway travel selection considering individual risk preference difference

Ying-ying MA1,2(),Si-yuan LU2(),Xiao-ming ZHANG2,Wen-shu WEI2   

  1. 1.School of Civil Engineering and Transportation,South China University of Technology,Guangzhou 510640,China
    2.Transportation Planning and Design Studio,Guangzhou Urban Planning & Design Survey Research Institute,Guangzhou 510060,China
  • Received:2020-06-11 Online:2021-09-01 Published:2021-09-16
  • Contact: Si-yuan LU E-mail:mayingying@scut.edu.cn;1055054928@qq.com

摘要:

本文旨在进行节假日高速公路出行特征分析及出行行为决策建模。首先,进行节假日高速公路出行行为特性调查和分析,总共回收有效问卷数量500份;其次,考虑个体风险偏好差异性,提出基于巢式Logit-累积前景理论模型的高速公路出行时段选择模型,进一步了解节假日高速公路出行行为决策;最后,以广东省节假日高速公路为对象进行案例研究,并进行模型精度检验。研究表明:本文模型总体预测误差为14.14%,仅基于巢式Logit模型预测误差约为22.22%。与其他模型相比,本文模型有利于提高预测结果的精度。

关键词: 高速公路, 巢式Logit模型, 累积前景理论, 风险偏好系数

Abstract:

This paper aims to analyze the travel characteristics of highway on holidays and establish the travel behavior decision-making model. First of all, a total of 500 valid questionnaires were collected from the survey and analysis of highway travel behavior characteristics on holidays. Secondly, considering the difference of individual risk preference, a freeway travel time selection model based on nested logit -cumulative prospect theory model is proposed to further understand highway travel behavior decision-making in holidays. Finally, a case study is carried out on the holiday Expressway in Guangdong Province, and the accuracy of the model is tested. The results show that the overall prediction error of the proposed model is 14.14%, and prediction error of the nested logit model is about 22.22%. Compared with other models, the proposed model can improve the accuracy of prediction results.

Key words: highway, nested Logit model, cumulative prospect theory, risk preference coefficient

中图分类号: 

  • U491

表1

调查者统计分布"

因素选项频率/%因素选项频率/%
年龄<25岁36.00月收入<5 000元34.80
25~35岁39.005 000~10 000元30.80
>35岁25.00>10 000元24.50
职业状态固定工作53.50教育水平高中及以下20.30
学生32.00大学生/专科45.50
其他14.50研究生及以上34.30
驾龄≤2年53.00小汽车拥有0辆40.80
2~4年25.801辆43.30
>4年21.30>2辆16.00

图1

收费节假日高速公路出行特征频率分布"

图2

收费节假日高速公路出发时段频率分布"

图3

主观效用(概率)与实际效用(概率)关系图"

表2

高速公路收费客车车型及额载人数"

车型额载人数
一类客车7座以下(含7座)
二类客车8~19座(含19座)
三类客车20~39座(含39座)
四类客车40座以上(含40座)

图4

基于出行时段-出行方式的巢式Logit模型结构图"

表3

参数φ取值列表"

调查规模/人φ取值调查规模/人φ取值
>10000.9~1.0300~5000.3~0.5
800~10000.7~0.9<3000.1~0.3
500~8000.5~0.7

表4

变量定义及取值"

因素变量取值变量名
个人及家庭经济属性性别男生取1,否则取2xgend
年龄1:<25岁;2:25~35岁;3:>35岁xage
月收入1:<5 000元;2:5 000~10 000元;3:>10 000元xincom
职业状态职业状态1固定工作取1,否则取2xjob1
职业状态2学生取1,否则取2xjob2
职业状态3其他工作取1,否则取2xjob3
教育水平教育水平1高中及以下取1,否则取2xedu1
教育水平2大学生、专科取1,否则取2xedu2
教育水平3研究生及以上取1,否则取2xedu3
驾龄1:<2年;2:2~4年;3:>4年xdage
小汽车拥有量1:0辆;2:1辆;3:2辆及以上xcnum
出行特征出行目的出行目的1哑变量,旅游休闲取1,否则取2xpur1
出行目的2哑变量,探亲访友取1,否则取2xpur2
出行目的3哑变量,其他取1,否则取2xpur3
陪同人数1:无;2:1人;3:2人;4:2人以上xcomp
出行耗时1:<1小时;2:1~2小时;3:2~4小时;4:>4小时xtime
乘坐车型1:小型客车;2:中巴;3:大巴xvehi
出发时段1:6∶00~9∶00;2:9∶00~12∶00;…;7:00∶00~06∶00xdepa
出行方案1:(时段1,车型1);…;21:(时段7,车型3)xdeci
舒适性1:差;2:中;3:好xcomf
便利性1:差;2:中;3:好xcomv

表5

不同收入水平不同出行方案的风险偏好系数"

出行方案(k)输入水平出行方案(k)输入水平
Y=1Y=2Y=3Y=1Y=2Y=3
10.930.910.95120.980.990.99
20.990.991.00130.960.970.98
30.980.980.98140.980.990.99
40.840.940.96150.990.990.98
50.980.990.99160.990.990.99
60.960.950.95170.990.980.96
70.930.940.96180.980.990.99
80.980.970.97190.990.980.97
90.980.990.99200.990.990.99
100.950.880.88210.980.990.99
110.980.991.00----

表6

出行者n累积前景值计算结果"

R组合效用价值函数累积权重CPV
1V1V2V3v(ΔV1)v(ΔV2)v(ΔV3)π1π2π3CPV1
0.31-4.56-0.440.34-10.24-1.010.200.0010.12-0.061
2V4V5V6v(ΔV4)v(ΔV5)v(ΔV6)π4π5π6CPV2
0.41-0.88-0.760.43-1.97-1.740.230.050.11-0.196
3V7V8V9v(ΔV7)v(ΔV8)v(ΔV9)π7π8π9CPV3
-0.33-0.878-2.41-0.75-1.98-5.380.110.040.01-0.195
4V10V11V12v(ΔV10)v(ΔV11)v(ΔV12)π10π11π12CPV4
0.21-2.98-1.630.26-6.66-3.650.280.010.09-0.358
5V13V14V15v(ΔV13)v(ΔV14)v(ΔV15)π13π14π15CPV5
0.24-3.14-1.930.25-7.06-4.340.190.010.04-0.215
6V16V17V18v(ΔV16)v(ΔV17)v(ΔV18)π16π17π18CPV6
0.21-0.75-30.830.22-4.56-61.590.140.07-0.13.491
7V19V20V21v(ΔV19)v(ΔV20)v(ΔV21)π19π20π21CPV7
0.05-42.860.350.05-95.940.350.060.000.04-0.061
实际决策出发时段:6=17∶00~21∶00;乘坐车型:1=小型客车

表7

不同出发时段主要影响因素的Od值"

出发时段xagexincomexjob3xedu1xdagexcnumxtime
R=10.6841.0032.7570.3140.8240.4970.436
R=20.5560.7761.6940.5261.1430.5050.501
R=30.5830.8971.6801.1341.0370.4240.643
R=40.8841.2200.7250.2250.9020.4150.512
R=50.6930.6980.9710.3860.6800.6560.519
R=60.4020.8861.0290.1980.7930.6000.534
R=7

表8

仅基于巢式Logit模型和基于巢式Logit-累积前景理论模型预测结果比较"

出发时段实际选择比例巢式Logit模型巢式Logit-累积 前景理论
预测结果误差预测结果误差
总体误差22.2214.14
R=16.066.060.007.071.01
R=237.3743.436.0631.316.06
R=315.1512.123.0315.150.00
R=425.2530.305.0526.261.01
R=54.042.022.023.031.01
R=66.062.024.0411.115.05
R=76.064.042.026.060.00
1 贾洪飞, 龚勃文, 宗芳. 交通方式选择的非集计模型及其应用[J]. 吉林大学学报: 工学版, 2007, 37(6): 1288-1293.
Jia Hong-fei, Gong Bo-wen, Zong Fang. Disaggregation model of traffic mode selection and its application[J]. Journal of Jilin University(Engineering and Technology Edition), 2007, 37(6): 1288-1293.
2 Kumagai S J, Uchiyama H, Takashima S. Formation of bus network based on traffic mode selection model with disaggregate data[J]. Physics Letters A, 2011, 375(17): 1831-1838.
3 Ozbay K, Yanmaz-Tuzel O. Valuation of travel time and departure time choice in the presence of time-of-day pricing[J]. Transportation Research Part A Policy & Practice, 2008, 42(4): 577-590.
4 张旭, 栾维新, 赵冰茹. 基于非集计模型的武广线高铁与民航竞争研究[J]. 交通运输系统工程与信息, 2012, 12(6): 17-21.
Zhang Xu, Luan Wei-xin, Zhao Bing-ru. Research on competition between Wuhan Guangzhou high speed railway and civil aviation based on disaggregation model[J]. Transportation System Engineering and Information, 2012, 12(6): 17-21.
5 McFadden D, Train K. Mixed MNL models for discrete response[J]. Journal of Applied Econometrics, 2000, 15(5): 447-470.
6 李华民, 黄海军. 基于一种新效用函数形式的分层Logit模型[J]. 吉林大学学报: 工学版, 2009, 39(): 63-65.
Li Hua-min, Huang Hai-jun. Layered logit model based on a new utility function form[J]. Journal of Jilin University(Engineering and Technology Edition), 2009, 39(Sup.2): 63-65.
7 唐洁, 隽志才, 高林杰. 城市居民出行空间和方式联合选择模型研究[J]. 公路交通科技, 2010, 27(5): 83-87.
Tang Jie, Zhi-cai Juan, Gao Lin-jie. Research on joint choice model of urban residents' travel space and mode[J]. Highway Transportation Technology, 2010, 27(5): 83-87.
8 诸葛承祥, 邵春福, 李霞, 等. 通勤者出行时间与出行方式选择行为研究[J]. 交通运输系统工程与信息, 2012, 12(2): 130-135.
Cheng-xiang Zhu-ge, Shao Chun-fu, Li Xia, et al. Study on commuter travel time and travel mode selection behavior[J]. Transportation System Engineering and Information, 2012, 12(2): 130-135.
9 陈林. 基于Logit模型的成渝通道交通选择行为研究[D]. 成都: 西南交通大学交通运输与物流学院, 2013.
Chen Lin. Study on traffic choice behavior of Chengdu Chongqing corridor based on logit model[D]. Chengdu: School of Transportation and Logistics, Southwest Jiaotong University, 2013.
10 Kahneman D, Tversky A. Prospect theory: an analysis of decision under risk[J]. Econometrica, 1979, 47(2): 263-292.
11 Kahneman A T. Advances in prospect theory: cumulative representation of uncertainty[J]. Journal of Risk & Uncertainty, 1992, 5(4): 297-323.
12 Avineri E J T S. The effect of reference point on stochastic network equilibrium[J]. Transportation Science, 2006, 40(4): 409-420.
13 施海燕, 施放. 期望效用理论与前景理论之比较[J]. 统计与决策, 2007(11): 22-24.
Shi Hai-yan, Shi Fang. Comparison of expected utility theory and prospect theory[J]. Statistics and Decision Making, 2007(11): 22-24.
14 刘玉印, 刘伟铭, 吴建伟. 基于累积前景理论的出行者路径选择模型[J]. 华南理工大学学报: 自然科学版, 2010, 38(7): 84-89.
Liu Yu-yin, Liu Wei-ming, Wu Jian-wei. Travel path selection model based on cumulative prospect theory[J]. Journal of South China University of Technology(Natural Science Edition), 2010, 38(7): 84-89.
15 Hu L, Dong J, Lin Z. Modeling charging behavior of battery electric vehicle drivers: a cumulative prospect theory based approach[J]. Transportation Research, 2019, 102: 474-489.
16 Guan Y, Annaswamy A M, Tseng H E. Cumulative prospect theory based dynamic pricing for shared mobility on demand services[C]∥2019 IEEE 58th Conference on Decision and Control(CDC), 2019: 04824.
17 Jhala K, Natarajan B, Pahwa A. Prospect theory-based active consumer behavior under variable electricity pricing[J]. Smart Grid, IEEE Transactions on, 2019, 10(3): 2809-2819.
18 邢睿. 基于前景理论的城市居民交通出行方式选择研究[D]. 长沙: 长沙理工大学交通运输工程学院, 2014.
Xing Rui. Research on urban residents' travel mode selection based on prospect theory[D]. Changsha: School of Transportation Engineering, Changsha University of Technology, 2014.
19 马书红, 周烨超, 张艳. 基于NL-累计前景理论的出行方式选择预测模型研究[J]. 交通运输系统工程与信息, 2019, 19(4): 135-142.
Ma Shu-hong, Zhou Ye-chao, Zhang Yan. Study on travel mode selection prediction model based on NL cumulative prospect theory[J]. Transportation System Engineering and Information, 2019, 19(4): 135-142.
20 杨志勇. 基于前景理论的出发时刻和出行路径选择模型研究[D]. 哈尔滨: 哈尔滨工业大学交通科学与工程学院, 2007.
Yang Zhi-yong. Research on departure time and travel path selection model based on prospect theory[D]. Harbin: School of Transportation Science and Engineering, Harbin Institute of Technology, 2007.
21 秦世环. 基于前景理论的出行方式与路径联合选择行为研究[D]. 哈尔滨: 哈尔滨工业大学交通科学与工程学院, 2013.
Qin Shi-huan. Research on joint choice behavior of travel mode and route based on prospect theory[D]. Harbin: School of Transportation Science and Engineering, Harbin Institute of Technology, 2013.
[1] 孙宗元, 方守恩. 高速公路出入口运动车辆轨迹分层聚类算法[J]. 吉林大学学报(工学版), 2017, 47(6): 1696-1702.
[2] 李烨, 王炜, 邢璐, 王昊, 董长印. 自动巡航与可变限速协同控制对高速公路基本路段通行效率的改善[J]. 吉林大学学报(工学版), 2017, 47(5): 1420-1425.
[3] 李鹏辉, 张文会, 胡孟夏, 李一兵, 吴彪. 高速公路驾驶人超车过程中的扫视行为[J]. 吉林大学学报(工学版), 2016, 46(1): 114-119.
[4] 杨庆芳, 马明辉, 杨聚芬, 邴其春, 梁士栋. 高速公路交通流运行稳定性分析[J]. 吉林大学学报(工学版), 2014, 44(6): 1596-1603.
[5] 詹伟, 吕庆, 尚岳全. 高速公路隧道群交通事故灰色马尔可夫预测[J]. 吉林大学学报(工学版), 2014, 44(01): 62-67.
[6] 徐铖铖, 刘攀, 王炜, 李志斌. 恶劣天气下高速公路实时事故风险预测模型[J]. 吉林大学学报(工学版), 2013, 43(01): 68-73.
[7] 王琳虹, 李世武, 周茹波, 杨志发, 冀秉魁, 姚雪萍. 高速公路路侧景观色彩对驾驶员心率指标的影响[J]. 吉林大学学报(工学版), 2013, 43(01): 74-80.
[8] 张文会, 许洪国. 高速公路交通事故现场路段限速方案[J]. 吉林大学学报(工学版), 2012, 42(02): 316-320.
[9] 徐亮,程国柱. 基于车速离散度和经济车速的高速公路最低车速限制[J]. 吉林大学学报(工学版), 2010, 40(03): 661-0665.
[10] 程国柱, 裴玉龙, 池利兵. 基于汽车行驶广义费用最小的高速公路最高车速限制方法[J]. 吉林大学学报(工学版), 2009, 39(04): 900-905.
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!