| Existing traffic theories and methods have made outstanding contributions in optimizing traffic system and serving residents' travel,but their foundation is relatively fixed physical space(road network structure)at the bottom.If the urban road network has the defect of restricting the traffic function,then a series of optimization measures may be difficult to solve the traffic problem.Therefore,it is necessary to study the relationship between the structure and function of urban road network from the perspective of traffic engineering.In the era of big data,urban traffic big data provides the possibility to comprehensively analyze the macroscopic characteristics of the urban road network and further empower urban traffic governance and service in the new era.This paper draws on network science and new urban science methodology to conduct data-driven research on the macro characteristics of urban road network,proposes research process and analysis method,and reveals the correlation characteristics of the underlying urban road network structure and function.It is innovative and practical,which can enrich the subject knowledge and provide theoretical and empirical guarantee for scientific and reasonable urban road network design and policy making.The characteristic of urban road network structure is the classical research content in the field of traffic network complexity.The second chapter summarizes the basic knowledge of the current research on the structural characteristics of the urban road network,including the form of the basic road network,the quantitative index of structural characteristics,and the method of topology reconstruction.Then the structural characteristics of the new road networks are analyzed,and the influence of road network density on motor vehicle travel is emphatically investigated.The third chapter links the structure and function of the road network with the lane attributes,focusing on the urban road network element connection characteristics.The degree index in complex network theory is transformed by lane attributes.The paper chooses the method of determining road network elements which accords with residents' cognition.The connection characteristics of road network elements are analyzed from three perspectives: intersection elements,road elements and road network skeleton.The results show that both intersection element network and road element network are scale-free networks.All the connections of road network elements are critical with a certain node showing segmentation characteristics.The degree distribution and connection characteristics of skeleton road network are consistent with the whole road network.The fourth chapter integrates the research ideas and technical tools of network theory,geographic information science,traffic engineering,etc.,and puts forward an analysis method of urban road network evolution,which mainly includes vector road network drawing,topological map generation,network index classification,etc.Among them,the index classification achieves the research goal of linking the two by classifying the state and functional indexes.The evolution of road networks in Changchun,China,a medium-sized developing city,is investigated.The results show that the evolution of urban road network is characterized by expansion and densification.The road networks of changchun in different periods are all small-world networks.Over time,the main roads in the network are consistent;however,the core areas are different,in adjacent areas and usually far away from the old city.More importantly,this chapter finds the correlation between structural indicators and functional indicators of the urban road network.This correlation relates to historical periods and planning patterns.Japanese plan(1932-1958)can be considered as the process of road network densification within the outer ring road,while China's plan(1968-2017)is the combination of densification and expansion.Compared with the first stage,the correlation increased between clustering coefficient,road length,efficiency and density of the second stage with other indicators.The fifth chapter puts forward the analysis process and method of spatial and temporal distribution of urban traffic demand and correlation with built environment based on traffic flow data and POI.Firstly,the temporal and spatial characteristics of traffic demand distribution are analyzed from two dimensions of time and space.In terms of time,the peak hour distribution and flow ratio of the three scales of the whole day,morning peak and evening peak are analyzed.In term of space,the paper studied the actual maximum service traffic volume of different roads.Then the built environment elements are extracted based on geographic detector,which provides the possibility to analyze the relationship between traffic flow and built environment elements.Finally,the SCATS and POI data of Shenyang city are used for empirical analysis.The results show that 88% of intersections and 93% of roads have a traffic flow ratio between 0.06 and 0.10,with an average traffic flow ratio of 0.08.There is a significant correlation between the peak hour flow and daily flow of road network elements.Traffic demand is mainly distributed in a relatively small number of roads.Medium and smallsized roads,which account for about 66% of the total roads,only bear about 38% of the traffic flow,while medium and large-sized roads,which account for about 34% of the total roads,bear about 62% of the traffic flow.In terms of the correlation between traffic flow and built environment,the influence of lane number is more significant than other factors.When the number of lanes at intersections and roads exceeds 14 and 4 respectively,lane resources are wasted to some extent.An improved gravity model based on simple structural parameters(lane number and road length)can be used to estimate daily and peak-hour flows in a given flow range.The sixth chapter puts forward an analysis method of urban residents' private car travel and path characteristics based on trajectory data and electronic map.Firstly,according to the definition of traffic travel,the rules of private car travel are defined.According to these rules,the travel routes of private cars are generated using geographic information technique,and the basic travel characteristics of private cars are described and analyzed.Then,based on the travel path of private cars,the shortest path(distance and time dimensions)data are collected by virtue of the path planning function of Baidu map to compare the difference and deviation between the travel path of private cars and the shortest path.Finally,the correlation between travel attributes and road network structure is analyzed.In this study,Chongqing results showed that a large proportion of private car trips did not choose the shortest path,and only 67% of trips chose the shortest path,significantly higher than foreign residents.No significant difference exists between workday and rest day.And the deviation between the actual path and the shortest path is small.The rest day was more skewed than the weekday.There are some differences in route selection and deviation in different travel modes,starting points,drivers' preferences,travel time,travel distance,etc.In addition to the investigation of the shortest path,residents' travel is correlated with the number of traffic lights at the intersection where the path passes,centrality index of nodes,travel directness,physical distance,road grade and direction change. |
PDF Full Text Request