Study On The Short-term Passenger Flow Forecast Of Urban Rail Transit

In recent years,urban rail transit has become more and more popular among travelers due to its convenience and other advantages.At the same time,excessive passenger demand has caused increased congestion and insufficient urban rail transit facilities.Therefore,it is particularly important to realize the scientific and reasonable planning of the departure interval of urban rail transit trains at different times,and to optimize the number and layout of facilities according to the passenger flow density of different stations.The urban rail transit short-term passenger flow forecasting can provide certain inspiration and effective data support for solving the above problems.As a key part of the designated future plan,accurate short-term passenger flow forecasting is of great significance for meeting passenger travel needs,improving the service level of operating departments and ensuring the economic benefits of operating companies.The paper focuses on the analysis and processing of urban rail transit short-term passenger flow data,the construction of urban rail transit short-term passenger flow prediction models,and the case analysis.First,according to the time-varying characteristics of urban rail transit passenger flow,the data is processed using algorithms such as clustering and decomposition.According to the daily passenger flow and the passenger flow by time period,the passenger flow data is clustered twice.At the same time,the EEMD and CEEMDAN algorithms are used to separate the data of each sub-cluster after the secondary clustering,which reduces the volatility of the original passenger flow data and ensures that the passenger flow data of each sub-cluster is more accurate in terms of prediction results.Then,combined with the characteristics of urban rail transit passenger flow,a short-term passenger flow prediction model that can accurately predict relevant data is developed.This paper proposes a support vector machine regression model based on gray wolf optimization algorithm to achieve accurate short-term passenger flow prediction.The gray wolf optimization algorithm has the advantages of fewer parameters and is not easy to fall into the local optimum,which makes up for the difficulty of finding a reference in the support vector machine regression model.At the same time,the BP neural network model optimized by the gray wolf algorithm is proposed,which greatly reduces the running time of the original BP neural network model.Finally,take Xi’an Metro Line 2 as an example to verify the feasibility of the model proposed in the paper,and compare the prediction results of related models.The hybrid forecasting model proposed in this paper is more effective,and the forecast results for nonlinear and non-stationary passenger flow data are more accurate.