Research On Spatial Distribution Prediction And Ground Access Mode Choice Of Large Airports In Cities

The airport is the distribution point of large passenger flow in the city.Due to its fast speed,convenient convenience,high comfort and wide radiation range,air transportation occupies an important position in the comprehensive transportation system.With the acceleration of China's economic development and urbanization process,air transportation has also developed to a new stage.The volume and scale of the airport are constantly rising.At present,the annual passenger throughput of large airports in China is generally more than 40 million.From the experience of construction planning of large airports at home and abroad,since large airports are often far away from urban areas,the introduction of rail transit has become more and more common.If there is no fast and convenient rail transit connecting airports and urban areas,the successful operation of large airports is difficult to achieve.Therefore,in order to ensure the speed and efficiency of urban air transport services,enhance the rationality of large airport planning and design,systematic research and analysis on a series of issues such as the demand distribution of large airports and passenger travel characteristics,has a very important theoretical value and realistic meaning.The paper relies on the sub-project 1(the strategic planning technology for large regional integrated transportation hubs based on airports)and sub-project 5(the superlarge airport coordinated rail transit layout planning and design technology)of the China Civil Aviation Administration's major special science and technology project(the research and application of key technologies for the construction of intelligent integrated transportation hub in Beijing New Airport),the research work takes large airports as the research object,respectively analyzes on the basic characteristics of large airport traffic,the spatial distribution forecast of large airport passengers,the choice of large airport passenger transportation modes,and spatial distribution prediction of large airports.The analysis of the basic characteristics of large airport traffic is based on multisource data such as field survey,mobile phone signaling and bus IC card.It proposes passenger travel OD identification method based on mobile phone signaling data and airport rail transit based on mobile phone signaling data.The passenger properties,travel characteristics,time and space distribution,bus travel characteristics,and passenger characteristics of the passengers in large airports are analyzed and summarized.The research on the spatial distribution prediction method of passengers in large airports is mainly based on the spatial distribution of passengers obtained by means of mobile phone signaling data.Considering the factors such as land use and population,the factors affecting the spatial distribution of passengers in large airports are analyzed.Then,it is determined that the resident population of the traffic community,the working population and the distance to the airport are input variables,and the percentage of passengers in the traffic community airport is taken as an output variable.A passenger urban spatial distribution prediction model based on modified gravity model,generalized regression neural network and genetic BP neural network is established,and the model prediction effect is compared and analyzed.By comparison,it is found that the effects of generalized regression neural network and genetic BP neural network prediction are better than gravity model,and the prediction effect of genetic BP neural network is better.Finally,based on the prediction method,the spatial distribution of passengers in the Capital International Airport in 2025 and the Beijing New Airport will be predicted.The research on the choice behavior of passenger transportation mode in large airports first analyzes the characteristics of airport passenger transportation mode selection and its influencing factors,and obtains the main travel chains of large airport passengers to and from the port.At the same time,based on the characteristics of passengers and the choice of transportation mode,the AP classification based passenger classification is proposed.Then,based on the travel chain,combined with the concept of generalized cost,the generalized cost function of the travel chain considering economy,rapidity,convenience,comfort,security and punctuality is established.And the large airport passenger transportation mode selection model based on the nonaggregate selection model is constructed for different types of passengers.Comparing the model with the model without considering passenger classification,it is found that the large airport passenger transportation mode selection model considering passenger classification has better prediction effect.Finally,taking Beijing New Airport as an example,the study on the proportion of travel modes of Beijing New Airport was carried out.The research on the spatial distribution prediction method of large airport rail passengers firstly analyzes the influencing factors of airport rail passenger flow according to the psychological process of passenger travel mode selection,and obtains two major influencing factors: the accessibility of large airport rail transit and the generalized cost of transportation mode.And the definition and model of the accessibility of large airport rail transit are proposed.Finally,based on the mobile phone signaling data,the generalized travel cost of various transportation modes and airport rail transit accessibility are used as input variables,and the airport rail passenger flow sharing rate in each traffic cell is used as the output variable to construct a generalized regression neural network model and genetics BP neural network model for prediction of airport rail passenger flow spatial distribution.Comparing the two models,it is found that the method of predicting the spatial distribution of passengers in large airport passenger airports based on generalized regression neural network is better.At the same time,taking Beijing New Airport as an example,the distribution of airport rail passengers at the new airport was predicted.Through the in-depth study on the basic traffic characteristics of large airport,the passenger spatial distribution forecast of large airport,the travel mode selection behavior of passengers in large airports,and the spatial distribution prediction of large airport rail passengers.The paper provides important scientific reference and theoretical support for the planning and construction of large airports in the future and the reasonable matching of urban related facilities.