Passenger Flow Prediction For Urban Railway Station’s Entrance And Exit Based On Land-use And Accessibility

With the continuous development of social economy, residents travel dramatic increase in demand, prompting urban rail transit network improves further and new lines are continuously constructed and put into operation. The increase of new stations add difficulty of prediction of urban railway station’s entrance and exit passenger flow.In this paper, a new approach of station’s entrance and exit passenger flow prediction include the existing line impact factor prediction model and the new line iterative feedback prediction model is proposed, which characterized with abandonment of detailed the economic data and land-use numerical indicators.First of all, land use is the source of passenger flow, traditional passenger flow prediction for urban railway station’s entrance and exit require land use numerical indicators (such as land area of the various categories, population, number of employees, shopping area, etc.), these data takes a lot of manpower, material and time. In the paper, Starting with mechanism analysis of trip generation, the influence factors on entrance and exit passenger flow are analyzed. Entrance and exit peak hour coefficients and whole-day entrance and exit passenger flow are selected as the input data.Sencondly, urban rail transit passenger flow closely related to the different traffic conditions, while the traditional entry and exit traffic predictions,which is a static prediction, only consider the impact of land use. Considering the advantage of infrastructure-based,location-based, person-based and utility-based measures on the basis of related literature at home and abroad, the paper proposes a new accessibility measure which is based on the utility function.Thirdly, the subway stations are classified by Fuzzy C-means including the fuzzy process of original data,the station accessibility and the construction of the fuzzy similarity matrixes of entrance and exit peak hour coefficients and whole-day entrance and exit passenger flow. Thus the input data (the station accessibility and the station land-use type and intensity) can match with the constructed historical database to predict the entrance and exit passenger flow roughly.And then, the existing station’passenger flow prediction for entrance and exit is on the basis of a comprehensive analysis of the impact of new lines to existing lines, and put forward five main impact situations with examples, and calculate the corresponding growth factors.Finally, the input data of accesibility can be modified by the accesibility measure based on the prediction of OD passenger flow to predict the entrance and exit passenger flow accurately.To evaluate the performance of the proposed approach, the entrance and exit passenger flow prediction of line8,line9and line15of Beijing operated on December31,2011is carried out, and the model is proved having a high accuracy as the rate of error is under10%.Conventional prediction model usually needs an accurate and detailed survey of socio-economic information (i.e. data indicates land-use type and intensity around stations), which takes a lot of time, manpower and financial resources. Compared with the conventional prediction model, the measure can be used in practice easily. The measure can be understanded and interpreted by researchers, planners and policy, thus it is likely to be used in studies of application of urban railway station’s entrance and exit passenger flow. The measure is sensitive to changes in the transport system and land-use system.