Research On Decision Mechanism Of Urban Household Members' Activity And Travel Behavior

The increasingly fierce contradiction between transport supply and travel demand has become a bottleneck restricting the sustainable development of the Chinese urban cities.However,simply increasing transport supply cannot fundamentally solve urban traffic problem.Therefore,traffic engineers began to seek new solutions,transforming the concept of urban transportation development into control of travel demand,that is,developing and implementing travel demand management strategies.In-depth study of urban residents' activity and travel behavior characteristics and rules is the key to formulating a scientific and effective travel demand management strategy.The study of the travel behavior involves the generation mechanism of the travel demand,the decision-making mechanism of the travel choice behavior,and the preference of the travel choice behavior,etc.Previous travel behavior research usually assumes that the travel choice behavior is individual,ignoring that a traveler is also a member of a household.Under the temporal and spatial restrictions,the travel behavior is not only driven by the needs of traveler's own activity demand,but also impact by factors such as other household members' activity and travel behavior,household structure,household resource allocation,etc.The study of the household-based urban residents' activity and travel behavior has shown rich achievement in some developed countries and regions,such as the data mining of the travel surveys,the influencing factors exploration,the analysis and modeling of the travel decisionmaking processes.However,this field is still at an initial stage in China,and there is no specific research on the differences in the characteristics of the Chinese urban household organization,urban land use structure and residents' activity and travel behavior.To this end,this paper is based on the data of the residents' trip survey,taking the household as the research unit,fully considering the social and cultural background and the characteristics of the urban household organizational structure,focusing on the similarities and differences of the behavioral characteristics of the household members with different identity attributes under the same temporal and spatial condition.The decision-making mechanism of the household members' activity and travel behavior is investigated from two levels,which are intro-household level and inter-household level.The specific research content has the following six aspects:(1)this paper focuses on the unique social and cultural characteristics in China.According to the theory of Hofstede's cultural dimension,it points out that “Power Orientation,Uncertainty Avoidance Orientation,Altruism Orientation,Masculinity Orientation and Long Term Orientation” are the main characteristics of the Chinese cultural behavior,which are different from other countries.This paper argues that,as an indispensable behavior in daily life,every household members' activity and travel behavior also reflects the cultural behavior characteristics.By expressing the activity-based travel demand analysis theory,it points out that household structure,the status of the individuals in the household,the importance and priority of activities,and the social-economic attributes of the household will affect the traveler's activities and travel decision-making behavior.Taking the characteristics of the Chinese household structure and the household registration system into consideration,this paper proposes that there are differences and interactions between the activity and travel behavior of the male and female household members,there are also differences and interactions between the household head and non-head household members'.(2)the data on the behavioral characteristics of the male and female household members,household head and non-head household members is collected and processed,taking advantage of the residents' trip survey.The paired t-test shows that there are significant differences between the activity and travel behavior of the male and female household members,household head and non-head household members,in terms of activity allocation,trip frequency,travel mode choice,etc.By using ArcMap software,through “geo-coding” analysis,various information about the built environment around the sampling households' residences is extracted,and the built environment characteristics are visually described.The data of activity and travel behavior is projected into the actual geographical environment,and the spatial trend analysis of the household members' activity and travel behavior is analyzed.It is concluded that there are differences in the activity and travel behavior of the household members living in different districts.(3)the structural equation model can simultaneously process multiple dependent variables and independent variables,and can flexibly formulate the directional relationship among variables.Based upon these advantages,the individual activity-travel behavior structural equation models of different household members are established,and the similarities and differences between different household members' individual activity and travel behavior interrelationship are compared.A structural equation model of household-based activity-travel behavior is also established,including household members' activity time allocation,travel duration interaction model,household members' trip chaining behavior interaction model and household members' travel mode choice interaction model.The model results prove that the activity and travel behavior of household members reflect the cultural characteristics of “Power Orientation,Masculinity Orientation,Long Term Orientation,and Altruism Orientation”,especially point out that under the unique Chinese household registration system,as the head of the household,household head's activity and travel behavior are different from other household members'.The model results enrich the household-based research on the urban residents' activity and travel behavior in China.(4)by presenting the characteristics of urban residents' travel behavior in Chinese cities,the significance of studying the impact of the urban built environment on residents' travel behavior is expounded.The two-step cluster analysis method is adopted to divide the built environment of the sampling household residence into three categories under the premise of considering the spatial correlation of the built environment characteristic.Through the design of matching the types of the sampling household built environment,using the propensity score matching algorithm,the actual impact of the built environment on urban residents' travel behavior is evaluated,meanwhile the heterogeneous impact on household members with different attributes is pointed out.Simultaneously,the residential self-selection effect is quantified,and the reasons for the underestimation of the built environment impact are analyzed from the perspectives of Chinese urban housing system and education resource distribution.According to these results,some strategies are proposed for the development of the city and transportation in this paper.(5)the definition of the urban residents' travel environment is extended and the influence of the neighborhood characteristics on the urban household members' travel mode choices is discussed by using spatial Durbin probit model.Model results show that female members are more sensitive to the choice behavior of the cars and active mode in the neighborhood,and will change the probability of choosing car and active travel mode.While male members are more concerned about their own activity and travel demand,and are less sensitive to external traffic conditions.In addition,Using the prediction function of the spatial probit model,the influence of the change of the household car ownership on the structure of the urban regional travel mode is analyzed.Because of the spatial correlation of the urban residents' travel mode choice,when the travel mode structure in a certain area changes,it will also change the structure of the travel mode in its adjacent area.(6)an urban residents' activity chaining pattern recognition and prediction model based on support vector machine is established,using personal attributes,household attributes and residential built environment characteristics.The feasibility and advantage of the support vector machine in the recognition and prediction of the activity chaining patterns is proved in terms of overall accuracy and partial accuracy,comparing with the multivariate discrete choice model and radial basis neural network model.Suggestions and directions for model improvement from the contribution of the explanatory variables are proposed.The research results in this dissertation promote the development of the research on the household-based activity and travel behavior in China,and enrich the content of the urban residents' travel demand analysis.In practice,this dissertation provides scientific and effective analysis tools for formulating urban transportation development strategies.