Division Method Of Urban Traffic Network Control Sub-regions Based On MFD And Evaluation Method Of The Importance Of Intersection On The Borders Between Sub-regions

The development of science and society have led to the dramatic population expansion in large and medium-sized cities and raised more travel demand among urban residents.At the same time,the spring up of electronic service providers have undoubtedly sped up the growing of logistics and transportation,which makes increasing severe shortage in traffic services.Nowadays,the congestion in major cities have turned out to be a serious problem around the world.In recent years,some foreign experts no longer focus on a single congested junction control strategy research,but focus on the basic nature of the entire road network,namely macroscopic fundamental diagram(MFD).The entire urban road network can be firstly divided into some homogeneous sub-regions in this method,and then the boundary control method is adopted between adjacent sub-regions in order to control the in-out traffic.The practice has proven the effectiveness of this method.As the research of MFD has just started in china,it has become a frontier problem in solving urban traffic congestion about how to design an effective control strategy of urban road network by exploring the MFD characteristics on the basis of existing transportation infrastructure.With using actual data of floating car in WangJing district(Beijing,China),the thesis puts forward a new method of city road network control sub-region division grid-by-layer based on MFD.At the same time,we have taken high efficiency of energy saving into account,this thesis proposes to exert boundary control only on key nodes along the sub-region boundary.Considering the basic characteristics of the general complex network and the topology of the actual road network,we propose a feasible evaluation method of the importance degree about the intersections,i.e.the evaluation of importance degree on m-order neighbor nodes.From simulation results and comparison with existing methods,the consequence proves that the algorithm has good feasibility and precision,which provides theoretical basis to realize the boundary control in road network nodes for future research.Firstly,this thesis carries on the pretreatment to the floating vehicle data collected in Wangjing area of Beijing city.Based on Gaode map,the road network is drawn out and the information of various intersections is counted.According to the GPS positioning data collected,the projection algorithm is used to match the road.On the basis of collected instantaneous speed of the floating vehicle,the required traffic parameters are calculated,such as:the total number of regional vehicles,the average speed,and the average density with the aid of vehicle identification,status and other data.Secondly,the traffic parameters are used to study the MFD characteristics of the whole road network.The research area is analyzed,and the region is divided by grid-by-layer trial and error method.The city road network MFD(macroscopic fundamental diagram)characteristics,such as uniformity,low hysteresis and discrete phenomena,are used as a series of evaluation indexes for the evaluation of the results of the regional division.Finally,the existing evaluation method of key nodes in complex network and the relevant road traffic index are analyzed.Then good characteristics embodied in the actual network are chosen out,such as degree,busy degree,closeness centrality and betweenness centrality.Moreover,a new evaluation method of key intersection based on the importance degree of m-level neighbor nodes is put forward.And then a new comprehensive evaluation index of the key intersections in the road network is established.The BP neural network based on the ant colony algorithm is used to evaluate the weights optimization.Compared with a variety of evaluation indicators,the feasibility and effectiveness of the proposed method are demonstrated.