The Research Of The Law Of The Terminal Departing Passengers Based On Improved K-Nearest Neighbor Algorithm

With the steady growth of China's civil aviation passenger traffic,China's major airports have varying degrees of operational capacity saturation.The traditional resource allocation methods of the terminal buildings can no longer meet the demand for airport passenger traffic growth.The major airports in the country have experienced a phenomenon of long queues for passengers and degraded passenger service quality.How to improve the utilization of airport resources under limited resources has become an urgent problem to be solved.The short-term passenger flow forecast of the terminal is the key core problem of the passenger service.The prediction accuracy directly affects the cracking effect of the above problems.It is found that the K-nearest neighbor(KNN)nonparametric regression algorithm is more suitable for the prediction of short-term passenger flow.However,because the traditional KNN algorithm does not consider the factors affecting the short-term passenger flow of the terminal,it has not good robustness.So we draw on the power system prediction method and introduce the "similar day" prediction idea.The flight schedule state pattern matching procedure is added on the basis of the traditional K-nearest neighbor algorithm.The flight schedule including multi-dimensional attributes is taken as a feature to select historical similar operation days as forecast reference vectors.The two-tier K-nearest neighbor model based on terminal building short-term passenger flow forecast is built(T-KNN).Experiments show that the prediction results of T-KNN model have higher precision and better robustness than the traditional KNN model.However,considering only one feature as an influencing factor does not have good persuasiveness.With the help of Airport Operation Control Center Big Data Platform,we further explore the characteristic attributes of the operation day,and add weather conditions,week types,and holiday impact factors on the premise of considering the flight plan impact factor.Comparing the two different days in the operation of the terminal building from the two aspects of similar characteristics and similar trends.Selecting the similar operation day of the forecast day,the SD-KNN model is established for the forecast of short-term passenger flow in the terminal.It is found that its relative T-KNN has higher precision and more stable robustness,which further improves the universality of the model.The simulation results show that the SD-KNN model better predict the short-term passenger flow of the terminal,with a prediction accuracy of about 93%.Successfully using the results in the airport evaluation project further verified the practical value of the model.