Terminal Departures Passenger Traffic Prediction Based On Chaotic Time Series

The improvement of people’s living standards tend to promote the traveldemand, airline occupies an increasing proportion in the traditional way to travelbecause of its fast and comfortable features. The growing traffic pressure on civilaviation field, inevitably led to terminal become congested, therefore, to achievehigh efficiency terminal operation put forward higher requirements on the arrivalterminal passenger traffic forecasts. For domestic research about prediction oftravelers arrive in a short time is still in its infancy, as the passenger traffic in a shorttime volatility, it is difficult using traditional linear methods to predict. Chaos theoryis a new research direction about nonlinear dynamical system, which is veryeffective for solving such complex problem of uncertainty. This article relies on theNational Natural Science Foundation project and chaos theory, for reference to otherareas of chaotic time series prediction research, using Matlab as an auxiliary tool,studied and analyzed the departures terminal passenger traffic.The paper first introduce the development of chaos theory and the basicconcepts, discuss and analyze the characteristics of chaotic motion and chaoticcharacteristic quantity calculation method; collected in the field for the time seriesdata of departing passengers arrival terminal at the airport. Using small-data methoddiscriminant the chaotic sexual. Through repeated calculations, we proveddepartures terminal passenger traffic at certain time intervals showing chaoticmotion characteristics.Secondly, reconstruct the phase space of chaotic departing passengers flow.Respectively, we calculated the chaotic characteristics such as optimal delay time,the best embedding dimension, correlation dimension. Then the paper using wavelettransform denoising method process the terminal departures passenger traffic andobtained the time series free of noise.Finally, this article analyses the steps, advantages and disadvantages ofdifferent chaotic time series prediction method comprehensively. Select the chaotictime series prediction method which is based on RBF neural network to establish ashort time terminal departures passenger traffic forecasting model. The size of themodel prediction error is validated by the collected historical passenger traffic data.In the end, we analyze the model’s meaning and application prospect in the terminal.