Study On Optimization Of Transit Station And Public Transportation Assignment Model During Peak Period

With the improvement of the society and the development of the economy, theprocess of our country’s urbanization and the development of the automobileindustry are both got effective propulsion. However, urban roads which extend in alldirections cannot ease the traffic congestion effectively. Traffic congestion,environment pollution and other problems have become increasingly serious. Notonly result in a waste of time, social idle labor force, but also greatly decrease thequality of people’s lives and affect the people’s health. Under the guidance ofsustainable development theory, it has become extremely urgent to advocate greentravel. On the one hand to promote public green travel and on the other hand to builda public transport system which meets the city’s needs, the latter is more importantobviously. As long as the public transport system can facilitate the people’s traveland offer better service, it can greatly increase the attractiveness of the passenger. Inthe planning of the public transportation network, whether public transportation canbe rationally designed, it has a big correlationship with whether the publictransportation can fully play its role. Currently, when we conduct the publictransportation planning, a “four-stage model” is well used for most cases, they aretrip generation, trip distribution, mode split and passenger distribution, and thepassenger distribution is the most important part.Morning and evening peaks are the most prominent time interval in travel jams,they are also the most dependent time interval for people’s daily lives in the publictransportation system. Whether we can find a more effective model to allocate thevolume of transportation in the peak time, it can result in directly whether publictransportation system can play a role in the green transport traffic problems in themost prominent period.In addition, the previous public transportation allocation models all lead to a bigdistribution problem which is influenced greatly by the shape of traffic zones, thecentroids of zones and relative position of the traffic zones, because the buspassengers are from the centroids of the traffic zones. If we can get the bus ODmatrix between each sites of the bus stations, and use it into the transport network,we can greatly reduce the impacts about the passengers’ allocation, which from theshapes of traffic zones and the centroids positions of the traffic zones, and it can alsoobtain a more reasonable results about the passenger distribution. What’s more,whether it conducts a reasonable planning about the bus stations, it not only affectsthe bus lines operation, but also affects the attractiveness of the passengers in the bussystem, as a result, the optimization about the layout of the bus stations is also a serious problem.Therefore, this paper relies on the project of the National Natural ScienceFoundation of China,“Research on public transit network generation optimizationmodel and algorithm based on network topology and traffic congestion (51378237)”,On the basis of domestic and international research, based on graph theory,Coulomb’s law, ultra-network theory and so forth, it focuses on the optimization thebus stations and bus passenger flow allocation in peak periods. With a starting pointabout bus station OD generations, we conducted bus station OD matrix’s generationbased on Coulomb’s law, rail station spacing optimization model and algorithm,routine bus station optimization models based on ultra-network and improvement ofthe Logit model about the allocation of bus passengers in the peak times. In thispaper, the research completed as follows：1. The study of the generation model about OD matrix between each busstation.Firstly, we analyze the characteristics about the attraction from the busstations to the passengers, with the introduction of Coulomb’s law, compared busstations to charged bodies, according to the spatial distribution of bus stations,computing attraction and repulsion between different stations, which were set up togenerate models which are OD matrix between each zone and OD matrix withinzones. According to the characteristics of different bus stations, optimized the modelwith an introduction of non-equilibrium factors. Finally, analyzed the sensitivitycoefficient of the model and achieved the results.2. The study of rail station spacing optimization.First of all, the shape of theouter contour of the proposed urban area was analyzed, the cities were divided intothree kinds of cities, and then the radius of the city was introduced to measure thegeographical size of the city. According to the traffic-related technical characteristicsof all levels of rail transports, lines were divided into two categories, established aspacing model about rail transit stations based on the average size of the cities, andthen got the results by analyzing the sensitivity coefficients. Then the lines weredivided into viable areas, hazard areas and connecting areas according to whether thestations can be built in the passing through area. Then according to thecharacteristics of the land and passenger volume in the passing through areas, thestations were divided into three levels, analyzed whether the ideal location of thestation is in a feasible region, and established a model which could determine thelocation of the railway station step by step.3. The study of regular bus station optimization.First of all, the traffic zone wassubdivided into source cells and defined the connections and nodes. Thenhypergraphs illustrate the relationship among different levels of roads, different sizesof source cells and different nodes. Source cells were divided into three levels byregion and population sizes respectively, after considering the region and populationsizes of the cell, the source cell size is divided into five levels according to the diagonal similarity of the matrix. The urban roads were divided into four levels andthe stations were divided into five levels, and subgraphs were drawn. According tothe hypergraph relationships among the source cells, urban roads and nodes,achieved the weights of different nodes. Lastly, optimized the stations on thesubgraphs step by step, and solved the weights of smaller cell sites.4. The study of the improvement of Logit model about passenger flowdistribution in the peak time. The degrees of crowdedness at stations were analyzedthen defined and solved the site retention factor. Then Logit model was improvedfrom time function, transfer penalty function and departure interval. Then LogitModels were established in terms of whether inter-sites have direct route or not andthe degree of crowdedness, stations were divided into direct line withoutcrowdedness, no direct line without crowdedness, and crowdedness stations. Finally,through the case study, analyzed the impact on the distribution of the results from theparameters αandT f, and marked the parameters ω�'�ωm axin the model.5. Relying on the research results from "Baicheng Public Transport Special Plan(2011-2030)" project, the model established in this paper was proved. Firstly,analyzed an overview of the2030White City urban planning and publictransportation network planning. Then optimized the light rail station’s layout. Thenthe nodes sub-graphs were drawn and the regular bus stations were optimized.According to the results of the optimization, distance matrixes were solved indifferent traffic zones, then achieved the correlation coefficient in OD generationmodel, and transformed the OD matrix within the traffic zones into the OD matrixwithin the bus stations. Combining with the situation through the bus line stations,the degrees of crowdedness at bus stations was assessed, and finally, the improvedLogit models were assigned to distribute the passenger flow and the distributionresults from the model and the distribution results from this project were compared.The results of this study enriched the theoretical system of public transportationplanning, provided a more accurate and solid data support for the public transitsystem, in the meantime, it has a higher theoretical and practical significance whileimproving the construction of urban public transportation system.