【作者】 栾琨；

【导师】 隽志才；

【作者基本信息】 吉林大学 ， 交通运输规划与管理， 2013， 博士

【摘要】 出行行为分析一直是交通科学领域研究的前沿课题，个体出行者在出行过程中的决策行为及其相互影响的结果表现为系统层面的流量、速度、可靠性等属性，对出行行为的分析和研究是诊断城市交通问题、制定交通管理政策的重要基础。路径选择和出发时间选择是出行行为的重要维度，为许多交通问题的研究提供了理论基础。传统的路径选择和出发时间选择行为的分析多是在规范假设下展开的，认为出行者是绝对理性的，面对临时决策时总有清晰的认识、明确的目标、无限的认知能力、稳定的偏好等等，并总是遵循效用最大化准则，较少考虑出行者的有限理性问题。本文围绕出行者的路径选择和出发时间选择问题，以出行者的出行决策行为研究为切入点，在有限理性的框架下构建基于规则的出行路径和出发时间选择行为模型，通过引入空间知识获取、学习及认知更新和方案搜索等关键行为要素，研究在认知局限约束下的出行决策行为。论文成果丰富了出行行为分析理论，为更合理、科学地制定交通需求管理政策提供理论指导和决策支持。论文进行了以下几个方面的研究：（1）以活动-出行理论为基础，分析了出行行为的三种决策理论。界定了有限理性决策过程的关键行为要素，解释了出行行为的形成和各关键要素之间的关系，为基于规则的出行行为分析提供了研究框架。（2）介绍了出行调查的两种方法：RP调查和SP调查，重点分析了SP调查的属性选择和情境设计原理，结合出行路径和出发时间选择行为对调查数据的具体要求，设计了基于序列正交设计法的出行路径和出发时间选择行为意愿调查方案。调查所得数据为后续研究提供数据支持。（3）将出行者的信息搜索、学习、认知更新和方案搜索等出行决策关键要素关联在一起，提出了基于RIPPER规则的出行路径搜索和选择行为模型。建模方法包括空间知识的呈现、搜索模型规范的定义、连续属性的离散化、属性的选择、路径搜索和选择规则的计算等。提出了主观搜索收益和感知搜索成本的概念，用以确定方案搜索开始和停止的条件。（4）构建了基于后悔最小化理论的出行者路径选择行为模型，详细分析了隐含在后悔最小化模型和效用最大化模型数学公式之内的决策机理，并结合出行路径的SP调查数据，分别建立采用完全理性和有限理性决策理论的路径选择模型，并分析了预测结果之间的差异。（5）将基于规则的出行行为分析框架用于出发时间选择行为的研究，提出基于PART决策树规则的出发时间行为模型，模型结合精心设计的出发时间调查方案，得到出行者出发时间搜索和选择的规则。论文的创新点主要体现在以下几个方面：（1）在有限理性框架下，提出了基于规则的出行路径和出发时间选择行为分析理论框架；在这个理论框架中，界定了有限理性决策过程的关键要素，包括出行预期、信息获取、学习和认知更新、方案生成和搜索过程，并将搜索理论应用在选择集的生成阶段；（2）定义了主观搜索收益和感知搜索成本，提出了基于出行预期的搜索过程启动和终止规则，构建了基于成本收益分析的出行选择方案搜索过程模型；（3）设计了基于序列正交设计法的出行路径和出发时间选择行为意愿调查方案，为基于规则的出行行为研究调查方案的设计提供研究基础；（4）分析了后悔最小化理论和效用最大化理论所蕴含的决策机理，解释了两个模型在决策过程和选择概率方面的差异。更多还原

【Abstract】 Travel behavior analysis is one of the frontiers of transportation science. Thesystem-level properties of transportation emerge from the behaviors of and theinteractions among a large number of individual decision makers. The researches on thetravel behavior analysis are the foundation of diagnosing urban traffic problems andmaking travel demand management policies. Route and departure time choice are theimportant dimensions of travel behavior. Most of the traditional route and departure timechoice behavior analyses are under the hypothesis of specific assumptions that traveler arecompletely rational, they always follow the utility maximization criterion, and seldomconsider bounded rationality in their decision-making.Based on the travel decision-making behavior as the point of research, this paper isintended to build the rule-based route and departure time choice behavior models in theframework of bounded rationality, and introduce the key subjective factors of behavior,such as spatial knowledge acquisition, learning, cognition updating, and alternativesearching. Travel decision-making behavior is analyzed under the restriction of cognitivelimitations. The main research contents are summarized as follows:(1) This paper defines and illustrates subjective factors related to the traveldecision-making process, and provides the research framework for the rule-based travelbehavior analysis.(2) Making comparative analysis between revealed preference and stated preferenceof travel behavior survey method, this paper focuses on the attribute selection andsituation design principle of SP survey. Combined with the specific requirements of routeand departure time choice behavior on survey data, the paper proposes and designs theroute and departure time choice behavior intention survey scheme based on the sequentialorthogonal design.(3) Making the subjective factors, such as the information search, learning, cognitiveupdating and alternative acquisition, related with each other, the paper proposes the route search and choice behavior models based on Ripper-rules. It also puts forward theconcepts of subjective search gain and perceived search cost, and uses these concepts todetermine the starting and ending conditions of alternative acquisition.(4) The paper constructs the route choice behavior model based on regretminimization theory, analyzes the underlying decision-making mechanism within themathematical formulas of regret minimization model and utility maximization model.Combined with the SP survey data of travel route, it also establishes the route choicemodel based on the perfect rationality and bounded rationality decision-making theoryrespectively, and analyzes the different between the estimation results.(5) Taking the rule-based travel behavior analysis framework as the basis foranalyzing departure time choice behavior, the paper proposes the rule-based departuretime choice behavior model, and builds the search and choice model of departure time.The main contributions of this dissertation research are as follows:(1) Under the framework of bounded rationality, this paper proposes the analysistheory framework of rule-based route and departure time choice behavior.(2) The concepts of subjective search gain and perceived search cost are defined, andthe starting and ending rules of expectation based search process are proposed. It alsobuilds the travel choice scheme search process model based on cost-benefit analysis.(3) The paper proposes and designs the route and departure time choice behaviorintention survey scheme based on the sequential orthogonal design, and provides researchfoundation for other rule-based travel behavior survey scheme.(4) The underlying decision-making mechanism within the regret minimizationmodel and utility maximization model are analyzed. The paper explains the difference indecision-making and choice probability.更多还原