【作者】 邵昀泓；

【导师】 王炜； 程琳；

【作者基本信息】 东南大学 ， 交通运输规划与管理， 2006， 博士

【摘要】 我国将大力推进国民经济发展和社会信息化作为现代化建设全局的战略举措。信息技术的飞速发展,影响了人们的活动决策过程,而传统的出行需求分析方法重点研究人和物的物理移动,没有考虑信息传递对交通运输的影响。因而在交通拥堵日益严重的背景下,本文对基于活动的出行需求分析及信息影响分析方法进行研究,具有重要的研究意义和应用价值。信息影响下的出行行为具有复杂性、随机性、动态性和特征依赖性,微观模拟方法根据调查统计的概率分布来复制人的行为,适用于本研究。本文的目的是结合常规居民出行调查数据,建立基于活动的出行行为模拟模型,并提出修正方法,研究信息对出行需求的影响。本文首先制定了信息提供条件下的出行需求分析框架,重点研究了基于活动的出行行为模拟方法及信息影响下的出行行为修正方法。接着采用“活动-出行模式”描述个体的活动参与和出行行为,建立了活动-出行行为模拟的概率模型,并应用蒙特卡罗方法按照时间顺序来模拟各特性变量。具体的模拟过程分为六个步骤:①按年龄和就业情况,将分析对象分组,并应用聚类分析法确定各分组的典型活动模式。②分析典型活动模式与社会经济特性的关系。③确定典型活动模式的概率分布。④确定出行特征变量的概率分布,是典型活动模式的条件概率分布。⑤模拟个体的活动-出行模式。⑥模拟结果的检验。考虑到活动模型和出行模型的分析对象不同,本文对数据选择和整理方法、有效性检查规则进行了详细阐述;应用沈阳市2004年的居民出行调查数据,分析初始活动-出行行为的特征,作为本文模拟的基础,并提出了基于交叉分类的出行量预测方法。然后制定了活动-出行行为的详细的模拟流程,研究了各特性变量的模拟方法;给出了模型的验证及确认方法,并编写计算机程序实现模拟流程;同时将提出的方法应用于沈阳市城市居民的活动-出行行为模拟中,将模拟结果与初始模式进行比较,结果显示误差很小,证明本文建立的模型正确合理,可用于信息影响研究。最后提出了信息影响下的出行行为修正方法——“增量分析”法。该方法模拟了信息影响下的个体行为反应,并在此基础上修正初始模式,确保其符合逻辑并满足时空约束条件。该修正方法可预测信息提供条件下的出行需求,对城市交通需求管理具有指导意义。更多还原

【Abstract】 Booting national economy and application of information technology (IT) have become an important strategic move for China’s modernization by the IT revolution. The individual’s activity decision process are affected by the rapid development of IT. The conventional method have focus on the physical movement of people and goods, neglecting the effects of information delivery on the transmission. In the context of increasingly severe traffic congestion, researching the activity-based travel demand analysis approach and method for the effect of IT on individual’s travel behaviour is of great significance.Information-affected travel behavior is characterized by complexity , randomicity, dynamics and reliant on attributes of individual. Micro-simulation attempts to replicate the individual behavior from a specific probability distribution, and is appropriate for our research. The purpose of the dissertation is to develop the activity-based micro-simulation model combining the trip survey data, and propose a modifying method to study the effects of information on travel demand.This dissertation established a travel demand framework with provided information, focusing on the activity-based travel behavior micro-simulation method and the modification approach for forecasting the travel demand affected by information. The activity and travel behaviour are depicted with“activity-travel”pattern and a model was established to simulate the activity characteristics with time step using the Monte Carlo method. The simulation program for activity-travel pattern includes six stages: Firstly, grouping all the individuals on age and employment status and classifying the individual activity-travel patterns as a number of representative patterns using the clustering methodology; Secondly, establishing the relationship between the identified representative patterns and individual socio-economic characteristics; Thirdly, estimating the representative patterns’choice probability distribution; Fourthly, given representative patterns specifying the conditional probability distribution of attributes for activity and travel behavior; Fifthly, simulating the individual activity and travel behavior; Sixthly, validating the simulated patterns.Considering the difference between the activity-based approach and trip-based method in research object, this dissertation expounds the trip data checking and the rules of validity judgement; With trip survey data of Sheng-Yang city in 2004, the character of initial activity-travel patterns are analized and a travel demand prediction method based on cross classification was put forward. And then, a micro-simulation flow for activity and travel behaviour was developed, studies the simulation model structure, proposes a validation method, programs to implement the simulation flow; applies the proposed model to the activity and travel analysis of Shen-Yang city; compares the simulation outcomes with actual patterns, results shows that the simulation errors is small. The proposed model has been proved to be correct and rational, and could be applied to the research of travel behavior provided information.The dissertation offered a modifying method for travel demand with provided information—incremental analysis. The approach simulates individual behaviour responses with provided information and modifies the initial patterns based upon the response in order to make it reasonable and satisfy time-space constraint condition. The modifyed approach could be used to forecast the travel demand with provided information and has significance for urban traffic demand.更多还原