The Reach Of The Law And Modeling Of The Terminal Departing Passengers

With the rapid development of China’s aviation industry, civil airport is facing increasing operating pressure, high rate of flight delays affects numbers of passengers. The slow response or improper disposal of airlines and airports makes passengers dissatisfied, which caused many incidents with bad influence; With the rapid growth in air passengers, China’s major hub airports are facing a severe test of saturated airport resources, which results in the long time in the queue to wait, which directly reduce the passenger travel experience. Hence building a passenger gathered warning measure and improving the utilization of airport resources become the current problems to be solved.Through analysis it is found that the accurate terminal departing passenger model becomes the key to solving these problems. To build the passenger model this paper starts from the law of single flight passengers. Through human dynamics analysis it is found that the behavior of single flight passengers in the group level has the characteristics of heavy-tailed distribution and mixed distribution. Combined with the time factor and regression analysis established Arrived Passenger Model in Single Flight based on the Time of Departure(TD-SFAPM). Considering the affect of departure time this model have good prediction accuracy. But for some special periods the accuracy of TD-SFAPM model is not good enough. Then this paper analyzed the error of TD-SFAPM and historical data. Through the establishment of the historical data evaluation, established the Improved Arrived Passenger model in Single flight based on Trust Level(TL-SFIAPM). The model makes up for the TD-SFAPM model and has a good accuracy in the special period. On this basis, established a timing rule then combined TD-SFAPM and TL-SFIAPM model. Through that method built the Model of Terminal Departing Passengers.Simulation result shows that this method can combine the number of departure flight and has a good prediction of the number of the departure passengers. The prediction accuracy can be above 80% and has a high application value.