Airport Passenger Throughput Forecasting Based On Improved Grey Model

As the development of civil aviation,the airport passenger throughput prediction is the focus of civil aviation business support and civil aviation construction.As a complex non-linear system,airport passenger throughput is affected by many internal and external factors,showing a certain regularity and continuous change trend.At present,the traditional forecasting methods generally have some shortcomings,such as low accuracy and limited time range.In order to predict the airport passenger throughput accurately,an improved grey model based on artificial neural network is proposed to predict the airport passenger throughput.Experimental results show that the air:port passenger throughput prediction based on the improved grey model has higher prediction accuracy.The main tasks of the dissertation are as follows:(1)Summarize the current research status of airport passenger throughput prediction,and introduce the relevant theories of airport passenger throughput prediction,including grey model and artificial neural network model.(2)This dissertation analyzes the relevant factors affecting airport passenger throughput prediction and the characteristics of airport passenger throughput changes in recent years,introduces the basic theory and algorithm of improved grey model(BP-GM(1,1))based on artificial neural network to correct residual error,and establishes the prediction model of airport passenger throughput based on BP-GM(1,1).(3)Experiments on the passenger throughput of Kunming Wujiaba Airport from 2000 to 2011 and Kunming Changshui Airport from 2012 to 2017 were carried out to compare the predicted results based on GM(1,1),BP neural network and BP-GM(1,1).The advantages and disadvantages of the above methods in the application of airport passenger throughput prediction were proposed.Based on the BP-GM(1,1),the validity of airport passenger throughput prediction is verified.