An Optimization Method Of Urban Rail Transit Local Network Based On Big Data

Urban traffic congestion and rapid motorized traffic become the first choice to deal with the problem of large urban traffic congestion and environmental protection.Although China’s urban rail transit started late,it has developed rapidly,and its operating mileage has ranked first in the world.While urban rail transit has made remarkable achievements,problems such as more passive transfer,long detour distance and insufficient attraction have gradually emerged.Take Beijing as an example.In 2020,the operating mileage of urban rail transit reached 799.4 kilometers,ranking the second in the world.However,the share of urban rail transit travel in the central urban area on weekdays is only 14.7%.How to effectively transform the existing network,optimize the connection relationship between the line network,improve the direct access of the rail network and enhance the attraction of urban rail transit has become an urgent problem to be solved under the background of the transformation from focus on construction to pay equal attention to construction and reformation in big cities such as Beijing.Based on this,this paper uses multi-source big data to start from three levels of "Area-Station-Line".Focusing on how to identify the urban rail transit rail network both attractive travel demand of weaker but short board area,how to choose a better alternative in the area of the short station,how to generate local line according to the selection of alternative station so as to optimize network relations,reduce passive transfer,promote urban rail transit line attraction effect were studied.The specific research work of this paper is as follows:Firstly,the travel characteristics of residents are analyzed based on multi-source data such as mobile signaling data and urban rail transit card swiping data.The detour coefficient index is constructed to represent the detour degree of urban rail transit passengers’ travel.And then,the areas with large amount of all mode travel between stations but high detour coefficient and few rail transit passengers in the existing rail network are identified with the help of detour coefficient and the difference between all mode and rail transit passenger flow,As a short board area covered by the existing network.Secondly,aiming at the problem of local line network optimization in short board area,firstly,this paper proposes a method to select potential rail transit stations and determine the route between alternative stations based on urban road network and conventional bus line network.Then,by systematically analyzing the influencing factors of line optimization in the feasibility study stage,and then aiming at the short board area with alternative stations,from the perspective of optimizing network topology structure such as improving network coverage and connectivity,the paper proposes three objective functions including passenger flow attraction,line engineering cost,line network average detour coefficient and six constraints such as station spacing under the existing network constraints.The single target line generation model and multi-target line generation model are constructed respectively.Then,how to solve the multi-objective optimization problem and the single objective optimization problem are described in detail.Combined with the professional and complex characteristics of the constructed route generation model,the classical non dominated sorting genetic algorithm with elite strategy and elite reservation genetic algorithm are modified,The non dominated sorting genetic algorithm with elite strategy for nested passenger flow allocation and the elite reservation genetic algorithm for nested passenger flow allocation are designed to solve the multi-objective line generation model and the single objective line generation model.Finally,taking the local short board area of Beijing’s existing rail network as an example,based on the potential rail transit station selection method proposed by the application of urban grade road network and conventional bus line network,this paper selects the alternative stations in the local area and the optional routes between stations,and then instantiates the line generation model by using the data of Beijing urban rail transit station spacing and so on.Finally,the instantiated model is solved by nested passenger flow allocation non dominated sorting genetic algorithm with elite strategy and nested passenger flow allocation genetic algorithm,and the new line scheme(s)under multi-objective line generation model and single objective line generation model are obtained.The results show that the generated local line scheme optimizes the connection relationship of the existing rail network and improves the direct access of the rail network.It can not only meet the regional travel demand and improve the attraction of urban rail transit,but also realize the better matching of the urban rail transit network topology with the actual passenger flow and urban spatial development layout,so as to provide a method for the fine-tuning and transformation of the rail transit network in the future.