Optimization Of Non-motorized Traffic Network Based On Complex Network Theory

As a major business activity district in Beijing,the CBD should be integrate with the international level.In order to create a more comfortable,safe and human-friendly non-motorized traffic(NMT)environment in the CBD,this dissertation studies the optimization of NMT network from the perspective of static traffic system and studies the scheduling problem of shared bicycle from the perspective of dynamic traffic.The main innovations of the thesis are summarized as follows:(1)Based on the complex network theory,the undirected and unweighted network model of NMT network in the CBD is constrycted by P space.The three indicators of degree,betweenness centrality and closeness centrality of the NMT network in P space are solved and clustered,whose results are used as the basis for the classification of NMT system.Then synthesizing the NMT flow and the road length as weights to construct an undirected weighted NMT network model by L space.In order to screen out the nodes which can be used as the key intersection and have critical impact on the robustness of the network,it is necessary to organize simulation experiments to conduct random attacks and deliberate attacks on the NMT network in L space.According to the distribution of key nodes on the network,it is judged whether the classification of NMT is reasonable.At the end,the strategy of improving the robustness of NMT network is proposed and the corresponding model is constructed.The model takes the robustness comprehensive threshold of NMT network as objective function and the cost of investment as the constraint.By solving the model,the edges or nodes that can be used as new paved NMT roads or intersections can be obtained,which constitutes the NMT network optimization scheme.(2)The data in this thesis mainly come from NMT travel demand questionnaire,on-site investigations and data uploaded by Mobike.The temporal and spatial variation characteristics of Mobike in the CBD are studied by analyzing the Mobike data.On the basis of complex network theory,the shared bicycle OD network model of the morning and evening peak period in the CBD is proposed by utilizing riding data of Mobike.Shared bicycle OD network in the CBD is analyzed by centrality index and cohesive subgroup analysis to solve the problem of shared bicycle scheduling center selection and scheduling area division.After determining the scheduling sub-area and scheduling center,the scheduling path planning model of shared bicycle is established.And then the circulation data of shared bicycle in a working day is selected as a case to path solution and demonstration analysis.Figure 59,table 19,reference 65.