Research On Quantitative Generation Model Of Urban Rail Transit Network

Urban rail transit network planning is an important stage in determining the shape,layout,and structure of the network.It affects not only the construction cost,travel efficiency and service level of the network,but also the form and sustainable development of the city.Due to intricate influencing factors,the formation of the network scheme is mainly based on qualitative analysis in practice.Some problems are gradually exposed in network operation in Beijing,Shanghai,Guangzhou,Shenzhen and other cities.The problems are mainly manifested in the incompatibility between the network and the passenger flow demand,resulting in the bottleneck for passenger flow in the network,overcrowding in some hub stations,and the normalization of flow restriction for stations.There are many reasons for the problems,but a very important reason is that the network-level problems in the network planning process have not been resolved.If the problems can be noticed and analyzed from the network level in the planning stage,some of them can be avoided in operation.Based on the understanding,the research on the quantitative model of network planning is carried out in this paper.In the paper,a quantitative model of urban rail transit network planning is established.In the model,the main characteristics and indexes of the city are taken as constraints;the trip attraction,the service performance of stations,the flow capacity reliability of paths in network,and the travel cost of passengers are taken as optimization objectives while the uncertainty of passenger flow demand and the adaptability of the network to the demand are considered.With the model,the automatic generation of the planning network under the influence of main factors is realized.The main contents and achievements in the paper are as follows:(1)The urban rail transit planning network should match the main characteristics of the city.In order to express the matching relationship in the model,two matching degree indexes are proposed: one indicates the matching degree between the network and the trip characteristics,which is calculated by the deviation between two gravity centers of the stations' importance distribution in network and the travel volume of residents;the other one describes the matching degree between the network and the land-use intensity,which is calculated by the deviation between the fractal dimensions of stations' importance distribution and the urban population distribution.In the “basic network” construction stage of the model,the trunk road network and the main locations of large traffic volume are used as constraint boundaries to express the matching relationship between the network and the main passenger flow corridors and main locations of large traffic volume.(2)The calculation methods of the indexes for the service performance of stations and the flow capacity reliability of paths in the network are put forward,so that the uncertainty of passenger demand can be taken into account in the network generation model,and the adaptation of the planned network to the future passenger demand can be improved.In view of the station service performance index in network,the predicted OD volume between stations is taken as the expected value of the future uncertain demand,and the OD random flow with the expected value as parameter is constructed.A passenger flow state evolution model with random flow input in stations is established.Based on the model,an algorithm for deriving the passenger flow status in stations is designed.With the model and algorithm,the queueing and stranding phenomenon of the passenger flow in stations can be deduced.Based on the results,the service performance index of the stations in network is calculated,and the applicability of the method is verified by an example.In view of the path flow capacity reliability index in network,a saturated flow loading model is established after the calculation of path selection probability.Based on the model,the flow capacity values for different saturated flow states of the network are calculated by simulating.Taking the path flow capacity as a random variable,the calculation method of path flow capacity reliability is put forward.The feasibility of the method is verified by calculating the basic networks.(3)A quantitative generation model of urban rail transit planning network is put forward.In the model,the maximum passenger attraction per unit length of the network,the minimum average generalized travel cost of passengers,the maximum average service performance of stations and the maximum reliability of flow capacity of paths in network are taken as objectives.The matching degrees between the network and the urban main characteristics,the network size,the anchoring stations,the service level,etc.,are taken as constraints.The coupling between the passenger demand and the network,and the uncertainty of future passenger demand are all well considered in the model by using the station service performance and path flow capacity reliability indexes.Because the solving of the model belongs to a NP-hard problem with high complexity and difficulty,a solution method based on hierarchical sequence and Pareto optimization is proposed.In the method,the candidate networks generation algorithm based on NSGA-II is designed,and the applicability of the algorithm is verified by numerical calculation.(4)The quantitative generation model of panning network is verified.The urban rail transit network planning in city X is taken as an example.Based on the model,the planning networks are finally generated.The results show that,the networks generated based on the model match the main characteristics of the city well.The networks have high passenger flow attractiveness and operation efficiency,and at the same time,they have good adaptability to future changes in passenger flow demand.The results of the paper provide a quantitative analysis method for urban rail transit network planning,which makes the network planning theory a step forward in the quantitative direction,and has a certain contribution to improve the rationality and scientificity of the network.