Prediction Of Security Check Passenger Flow In Terminal Building Based On Time Series Analysis

The improvement of people's living standards will often promote the demand for travel,and aviation,with its fast and comfortable characteristics,occupies an increasing proportion in the traditional way of travel.The increasing traffic pressure in the field of civil aviation inevitably leads to congestion in the security check of terminals.Therefore,in order to achieve efficient operation of the security check of terminals,it is necessary to put forward higher standards for the security check passenger flow prediction.If we can predict the security check passenger flow accurately,we can dynamically open security channel number and configure security service personnels according to the passenger flow,in order to ensure that the majority of passengers can complete the security check program in a short period of time,also reduce the possibility of the flight delays and gathering incidents caused by passengers' long waiting,and improve the service level,resource utilization on the ground,and the airport passenger satisfaction.Therefore,the study of the security check passenger flow prediction of the terminals method has important significance.Firstly,this paper summarizes the research background,significance and current research status of passenger flow prediction at home and abroad,analyzes the shortcomings of the prediction methods and models,and the good prediction effect of chaos theory in other fields,and adds chaos theory to the security check passenger flow prediction in this paper.Then,on the basis of studying the chaotic characteristics of the security check passenger flow,this paper presents the hidden information in the one-dimensional time series of the security check passenger flow through the phase space reconstruction technology,so as to better characterize the characteristics of chaotic attractors.In this paper,using the method of autocorrelation function to calculate time delay needed by the phase space reconstruction and adopting the false nearest neighbor method to calculate embedding dimension,and after reconstructing phase space of the real security check passenger flow data in Beijing capital international airport T3 terminal,using Wolf method to calculate the maximum Lyapunov exponent,and considering different time scales,such as 2 min,5 min and 10 min,the calculation results are positive,so the time series have chaotic characteristic.On basis of it,C-C method is introduced to improve phase space reconstruction method to strengthen calculation performance and decrease embedded dimension.The experimental results based on the real security check passenger flow data in Beijing capital international airport T3 terminal also testify that C-C method is effective in phase space reconstruction.Finally,the security check passenger flow prediction model based on the BP is regarded as a benchmark,and the security passenger flow prediction model based on GABP is proposed,considering different time scales(2 min,5 min and 10 min),with relative error for forecasting evaluation indicator,based on the real security check passenger flow data of the T3 terminal building of the Beijing capital airport,and the experimental results show that the prediction effect of the 2min time scale using GABP prediction method is better,and the relative error is smaller;Adopting C-C method to reconstruct phase space can not only decrease computational cost,but also improve prediction accuracy at some extent.