深圳大学学报理工版

目前制定交通政策考虑的因素不尽完善,导致政策实际效果与预期有差距. 为更好评估交通政策变化对消费者福利(出行者选择效益)的影响,更有效捕捉出行方案之间的相似特点,基于后悔理论,联合出发时间、出行方式和路线3个维度构建随机后悔广义巢式logit模型. 结合案例验证,在该模型结构下,使用Logsum消费者剩余度量方法评估政策变化对出行者效益的影响. 结果表明,当一个巢的价格属性或非价格属性变化时,该巢内部对应方案的消费者剩余都会呈不同程度的反比例变化,该方法能较好捕捉方案之间的相似性. 研究结果可为相关部门制定和评价交通政策提供依据.

At present, the factors considered in formulating the transportation policy are not complete, which leads to the gap between the actual effect and expectation. In order to better evaluate the impact of transportation policy changes on consumer welfare(i.e. traveler's choice benefit), and to more effectively capture the similar characteristics among travel schemes, we develop a generalized nested logit model of random regret based on regret theory which is constructed by combining three aspects of data: the departure time, travel mode and route. In case study, the selection probability is interpreted as an approximate probability demand curve, and the Logsum consumer surplus measurement method is used to evaluate the impact of changes between two policies on traveler benefits under this model structure. The result shows that when the price attribute or the non-price attribute of a nest changes, the consumer surplus of the corresponding alternatives within the nest changes in varying degrees in inverse proportion, thus it better captures the similarity between the schemes. The research results provide reference for relevant departments to formulate and evaluate transportation policies.

引言
1 随机后悔广义巢式logit模型的构建
2 基于随机后悔GNL模型的出行效益评估方法
3 案例分析
4 结论

图1 GNL模型的出行选择结构<br/>Fig.1 The travel choice structure of GNL model

图1 GNL模型的出行选择结构
Fig.1 The travel choice structure of GNL model

表1 出行选择方案组合<br/>Table 1 A combination of travel options

表1 出行选择方案组合
Table 1 A combination of travel options

图2 单一属性的消费者剩余变化<br/>Fig.2 Consumer surplus change with single attribute change

图2 单一属性的消费者剩余变化
Fig.2 Consumer surplus change with single attribute change

图3 价格为0时两种非价格属性情况下的消费者剩余变化

表2 以早高峰为例的出行者选择方案场景<br/>Table 2 Scenario description of traveler's choice scheme taking morning rush hour as an example

表2 以早高峰为例的出行者选择方案场景
Table 2 Scenario description of traveler's choice scheme taking morning rush hour as an example

表3 模型的变量<br/>Table 3 The variables of model

表3 模型的变量
Table 3 The variables of model

表4 随机后悔GNL模型的参数估计结果<br/>Table 4 The parameter estimation results of the random regret GNL model

表4 随机后悔GNL模型的参数估计结果
Table 4 The parameter estimation results of the random regret GNL model

表5 公共交通价格属性变化的Logsum衡量结果<br/>Table 5 Logsum measurement results of price attribute changes

表5 公共交通价格属性变化的Logsum衡量结果
Table 5 Logsum measurement results of price attribute changes

表6 非价格属性变化的Logsum方法衡量结果<br/>Table 6 Logsum measurement results of non-price attribute changes

表6 非价格属性变化的Logsum方法衡量结果
Table 6 Logsum measurement results of non-price attribute changes

[1] JONG de G, DALY A, PIETERS M, et al. The logsum as an evaluation measure: review of the literature and new results[J]. Transportation Research Part A: Policy and Practice, 2007, 41(9): 874-889.
[2] 马书红, 汤薛艳, 叶建勇. 基于Logsum差异和嵌套Logit模型的交通出行者效益测算研究[J]. 交通运输系统工程与信息, 2016, 16(5): 45-50, 78.
[3] MCCONNELL K E. Consumer surplus from discrete choice models[J]. Journal of Environmental Economics and Management, 1995, 29: 263-270.
[4] DEKKER T, CHORUS C G. Consumer surplus for random regret minimisation models[J]. Journal of Environmental Economics and Policy, 2018, 7(3): 269-286.
[5] STANDEN C, GREAVES S, COLLINS A T, et al. The value of slow travel: economic appraisal of cycling projects using the logsum measure of consumer surplus[J]. Transportation Research Part A: Policy and Practice, 2019, 123: 255-268.
[6] CHEN A, PRAVINVONGVUTH S,XU Xiangdong, et al. Examining the scaling effect and overlapping problem[J]. Transportation Research Part A: Policy and Practice, 2012, 46(8): 1343-1358.
[7] WEN C,KOPPELMAN F S. The generalized nested logit model[J]. Transportation Research Part B: Methodological, 2001, 35: 627- 641.
[8] 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.
[9] 杨励雅, 李娟. 居民出行链、出行方式与出发时间联合选择的交叉巢式Logit模型[J]. 北京大学学报自然科学版, 2017, 53(4): 722-730.
[10] 李雪飞. 基于扩展Logit的交通分配模型与算法研究[D]. 北京: 北京交通大学, 2014.
[11] CHORUS C, CRANENBURGH S, DEKKER T. Random regret minimization for consumer choice modeling: assessment of empirical evidence[J]. Journal of Business Research, 2014, 67(11): 2428-2436.
[12] CRANENBURGH S V, CHORUS C G. Does the decision rule matter for large-scale transport models[J]. Transportation Research Part A: Policy and Practice, 2018,114: 338-353.
[13] 李梦, 黄海军. 后悔视角下的多用户多准则随机用户均衡模型[J]. 系统工程理论与实践, 2017, 37(5): 1322-1330.
[14] GUEVARA C A, FUKUSHI M. Modeling the decoy effect with context-RUM models: diagrammatic analysis and empirical evidence from route choice SP and mode choice RP case studies[J]. Transportation Research Part B: Methodological, 2016, 93: 318-337.
[15] BIERLAIRE M. BIOGEME: a free package for the estimation of discrete choice models[C]// The Third Swiss Transportation Research Conference. Ascona, Switzerland: EPFL, 2003: 1-24.

备注

引言

1 随机后悔广义巢式logit模型的构建

2 基于随机后悔GNL模型的出行效益评估方法

3 案例分析

4 结论

期刊信息

备注

引言

1 随机后悔广义巢式logit模型的构建

2 基于随机后悔GNL模型的出行效益评估方法

3 案例分析

4 结 论

期刊信息

4 结论