基于差异化分类的出行联合决策模型研究

发布时间：2018-10-10 07:53

【摘要】：近年来,城市社会经济的快速发展及机动车保有量的不断增加,导致行车难、停车难等问题日益突显,交通拥堵问题已经给人们生活带来了极大的不便。交通拥堵现象是交通流在特定时间、空间下所表现的一种状态,其本质是出行需求主体——大量个体出行者出行行为决策实施宏观涌现的结果。目前的研究大多建立在出行者同质的基础上,忽略了出行者之间的差异性;且集中在单方面的决策行为上,对于两种及两种以上决策行为的联合决策研究较少。由此可见,有必要对通勤出行下的出行联合决策行为进行研究,分析不同出行群体的出行联合决策行为,为制定精细化的交通需求管理策略提供科学依据。首先,在分析出行决策影响因素的基础上,阐述了出行者的异质性,并对出行者差异化分类的方法进行对比分析;通过行为调查和意向调查获取基础调查数据,在对调查数据的有效性和可靠性进行检验的基础上,对出行者的社会经济属性进行统计分析,同时利用聚类分析方法对出行者进行人群归并,形成出行者差异化的分类结果。其次,根据出行者差异化的分类结果,对不同出行群体的出行特征进行归纳分析,采用主成分分析法确定通勤出行下不同出行群体的参照点构成因素为行程时间和到达时间;利用不确定理论构建行程时间及到达时间的参照点函数,并利用参照点及风险偏好系数的联系,构建风险偏好系数函数,同时结合贝叶斯更新模型对出行前及出行后的参照点及风险偏好系数函数进行更新。然后,从出行行为形成及响应过程剖析出行选择行为决策机理,并对出行联合决策进行界定;根据出行者的联合决策过程,选取前景理论对通勤出行联合决策进行分析;建立用于描述各备选方案可能发生结果的价值函数、感知概率函数及权重函数,将计算所得的价值函数、感知概率函数及权重函数带入计算公式中,计算得到各备选方案的前景值。最后,选取西安市的路网作为实例,调查得到路网的基础数据,在出行者差异化分类的基础上,构建基于前景理论的出行联合决策行为模型,分析在通勤出行时不同出行群体的出行联合决策行为,结果表明建立的出行决策模型与实际结果较为符合,可以准确地描述不同类型出行群体的出行决策行为。
[Abstract]:In recent years, with the rapid development of urban social economy and the increasing number of motor vehicles, the problems such as driving difficulties and parking difficulties have become increasingly prominent. Traffic congestion has brought great inconvenience to people's lives. Traffic congestion is a state of traffic flow in a specific time and space, and its essence is the result of macro emergence of travel behavior decision of a large number of individual travelers. Most of the current studies are based on the homogeneity of travelers, ignoring the differences between travelers, and focused on unilateral decision-making behavior, and for two or more kinds of decision-making behavior joint decision-making research is less. Therefore, it is necessary to study the travel joint decision-making behavior under commuter travel, analyze the travel joint decision-making behavior of different travel groups, and provide scientific basis for the formulation of fine traffic demand management strategy. First of all, on the basis of analyzing the influencing factors of travel decision, this paper expounds the heterogeneity of travelers, and compares the methods of passenger differentiation classification; obtains the basic survey data through behavior survey and intention survey. On the basis of testing the validity and reliability of the survey data, this paper makes a statistical analysis on the social and economic attributes of the travelers, and at the same time uses the method of cluster analysis to merge the travelers and form the differentiated classification results of the travelers. Secondly, according to the classification results of different travel groups, the travel characteristics of different travel groups are summarized and analyzed. The principal component analysis method is used to determine that the reference points of different travel groups under commuting travel are travel time and arrival time. The reference point function of travel time and arrival time is constructed by using uncertainty theory, and the risk preference coefficient function is constructed by using the relation between reference point and risk preference coefficient. At the same time, the reference points and risk preference coefficient functions before and after travel are updated with Bayesian updating model. Then, the decision-making mechanism of trip choice behavior is analyzed from the process of trip behavior formation and response, and the travel joint decision is defined. According to the joint decision process of travelers, the prospect theory is selected to analyze the joint decision making of commuting trip. The value function, perceptual probability function and weight function are established to describe the possible results of each alternative scheme, and the calculated value function, perceptual probability function and weight function are introduced into the calculation formula. The foreground value of each alternative is calculated. Finally, the road network in Xi'an is selected as an example, and the basic data of the road network are investigated, and based on the differentiation classification of travelers, a travel joint decision behavior model based on prospect theory is constructed. The travel decision behavior of different travel groups during commuting is analyzed. The results show that the established travel decision model is in good agreement with the actual results and can accurately describe the travel decision behavior of different types of travel groups.
【学位授予单位】：长安大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：U491

【参考文献】