Research Of Urban Traffic Network Complexity Measurement

The composition scale and combination way of urban traffic network is large and varied,consequently the urban traffic network is a complex covering urban geographical spatial system.In terms of morphological distribution and topology,it shows obvious complexity.The complexity of morphological distribution affects the overall planning of traffic network,additionally,the complexity of topology and the local traffic network depend on each other.Hence,the complexity of urban traffic network research has been widely concerned.However,the most existing research were applied a single method to measure the network complexity.In allusion to the different traffic data and network model,it is unclear to select a appropriate measurement method.Therefore,in this paper,three measurement methods were comparably analyzed,including fractal theory,complex network theory and space syntax.In addition,the characteristics and applicability of three methods were explored,the related technologies of different methods were studied.All of these can provide a basis for choosing the appropriate method,and lay a foundation of improving the efficiency of traffic planning.After analyzing the similarities and differences of the three methods,this article researched the building technology of urban traffic network model,studied the ArcGIS Engine secondary development technique and visual c#programming language,developed a software for calculating various kinds of complexity metrics.After that,the text taken the Nanjing traffic data as an example,analyzed the complexity characteristics of Nanjing transport network,further clarified the applicability of each measurement method.The main research results are listed:(1)The fractal theory is suitable for tremendous data,and it mainly reflects the shape distribution characteristics of traffic network.Besides,the network model can not consider build topology relationship when adopting fractal theory.The complex network theory is also appropriate for the gigantic data,and it mainly displays the topology and function of the network.And the network model must build topology relationship while choosing the complex network theory.Furthermore,when the traffic data can meet the requirements of space division,we may select the method of space syntax,which mainly analyzes certain network functions,involving the accessibility of local network,the prudent of overall network and the relevance of overall network.(2)Utilized three methods to measure the complexity of Nanjing traffic network,the verification results showed that the Nanjing road transport network has obvious fractal characters.The Nanjing road network revealed superior coverage degree and connectivity degree,but the cover depth needs to be improved.Additionally,the distribution of Nanjing regular bus routes and bus stops both have fractal characters.Nonetheless,the connectivity between bus stops is poor,the repeat rate of bus lines is very high,and the connectivity between each bus line is unreasonable.Moreover,the size of the Nanjing rail transit network is small,the accessibility between rail lines is unqualified,the network efficiency is low,and all the metro transfer stations are easy to be a congested area.(3)In the process of researching three measurement methods,this paper discovered that the branch of dimension can be used to measure fractal characteristics of bus stops.It means that the fractal theory is not only suitable for linear elements,but also can be applied to study the characteristic of shape distribution of point elements.Besides,we found that complex networks theory just focuses on the connectivity and reliability between the bus stops when utilizing the bus station network model.And it just emphasis on the connectivity and accessibility between the bus lines when employing the bus lines network model.Furthermore,for road traffic network,space syntax can apply axis method to spatial segmentation.For conventional bus network and rail transit networks,the routes can be regarded as point and the connectivity between lines can be regarded as line in the network.