Research On Passenger Volume Forecast Of Beijing Civil Aviation Based On Combination Model

With the development of social economy in our country,more and more people are pursuing efficient and comfortable means of transportation.Therefore,in recent years,China's civil aviation industry has made great progress in both scale and service quality.The passenger traffic volume of civil aviation is not only a reference to reflect the busy degree of civil aviation transportation industry,but also a main index to formulate transportation production plan and research transportation development.Therefore,how to use scientific methods to predict the passenger volume becomes particularly important.With the operation of Beijing Daxing International Airport in September 2019,it will have a huge impact on Beijing Civil Aviation Industry and become more practical for the prediction of Beijing civil aviation passenger volume.Based on this,this paper chooses to predict the passenger volume of Beijing civil aviation.Firstly,the paper constructs the influencing factor index system of Beijing civil aviation passenger traffic volume from the four aspects of economy,tourism,competition and airport operation ability,and finally establishes five main influencing factors by using correlation analysis method;then,it constructs ARIMA model,BP neural network model,series combination model and parallel combination model by using training data to improve Beijing civil aviation passenger traffic volume Line prediction,and use test data to verify the prediction accuracy of the model;finally,build a model evaluation system to measure the model.The results show that the ordered weighted combination model of the parallel combination model has the best effect in the prediction of Beijing civil aviation passenger traffic volume.Considering the fact that the parallel ordered weighted combination model can not predict the future passenger traffic volume,this paper proposes an improved parallel combination model to predict the future passenger traffic volume of Beijing civil aviation Measurement.The accuracy of the first mock exam is verified by the accuracy of the result.The empirical analysis shows that the accuracy of each prediction model is from high to low in the order of parallel combination model,ARIMA model,series combination model and BP neural network model.Among them,the ordered weighted combination model of parallel combination has the highest prediction accuracy,with an average relative error of only 1.01%;the BP neural network model has the lowest prediction accuracy,with an average relative error of 5.13%.The average relative error of the improved model is 1.75%,which is 1.02% lower than ARIMA model,and nearly twice as high as that of the series model.The improved parallel combination model is used to predict the passenger volume of Beijing civil aviation.According to the prediction results,the passenger volume of Beijing civil aviation will continue to grow steadily in the next 12 months,and will not fluctuate greatly compared with the past 36 months.