Prediction And Analysis On Terminal Departure Passenger Flow Based On Big Data

As the demand for civil aviation increases year by year,due to the high daily dynamic flow of passengers entering and leaving the airport and the uneven characteristics in time and space,the high-intensity and full-load operations of large and medium-sized airport terminals are not rare,peak congestion and even chaos occur frequently during the period,and low efficiency causes a series of problems and criticisms.Therefore,the development of the airport in the direction of "digitalization" and "intelligence" is an inevitable trend.In the systematic planning and decision-making process of smart airports,airport passenger flow forecasting is particularly important as a basic support.Accurate long-term forecast of airport passenger flow can provide reference basis for airport managers to optimize flight organization,airport terminal and airport runway expansion;accurate short-term passenger flow forecast of airport can scientifically organize check-in and security inspection,intelligent dynamic allocation Airport terminal resources,thereby reducing airport operating costs,improving airport operating efficiency,and improving airport service levels and passenger satisfaction.So,the accurate prediction of airport passenger flow is of vital importance.This paper first analyzes the research status of airport passenger flow prediction,preprocesses the data obtained from the Hohhot airport passenger flow,analyzes the factors affecting airport passenger flow,and divides the passenger flow status intervalThen,the use of airport passenger long-term traffic data and related influencing factors to establish a long-term forecast model of airport passenger traffic,that is,the forecast and application analysis of airport passenger traffic in a certain month or a certain future year.Using the error as an evaluation indicator,a comparative analysis of the established forecasting model is conducted,and it is found that the prediction based on the Prophet model is the best in airport long-term passenger forecasting,with the smallest error.Finally,combined with the historical passenger short-term flow data of the airport and related influencing factors,an airport passenger flow short-term prediction model was constructed,that is,the airport passenger flow in the next hour was predicted,and through parameter adjustment,the passenger flow short-term was finally obtained forecast result By comparing with different prediction models,the results show that the Gated Recurrent Unit(GRU)model is more suitable for short-term prediction of airport passenger flowThis paper establishes a long-term prediction model and a short-term prediction model for airport passenger flow.The results of simulation experiments show that the Prophet model and GRU model have the characteristics of high fit and small error in long-term and short-term prediction respectively,which can be used for the management of current smart airports and construction to provide a scientific basis for reference.