Ga - Svm Research And Application In Aviation Passenger Volume Forecast

With the improvement of social development and people’s material level, it is more often and common for people to travel by airplane. The global volume of air passenger transportation will rise by4.8percent per year. Facing the yearly growth of the volume of passenger travel on the aircrafts, if the volume of passenger is predicted in advance, it can be used as the decision-making basis of the flight route planning and crew scheduling in the process of management and operation.The accurate degree that the airline company predicts the volume of passenger influences even dominates the direction of the development of the airline as well as the accuracy of the airlines’decision-making. The prediction method is particularly important due to the randomness and time variability of the passenger traffic. The shortcoming of the traditional methods is that they can not predict the change trend and results of the passenger flow, that is to say, they have accurate prediction,provided that the change trend of the passenger flow data statistics don’t have big abnormal fluctuations or the statistical data are more than enough.The forecast accuracy may be affected as a result of the subjectivity and prediction for the related parameter,so the traditional method can’t meet the requirements of the airlines well.In order to solve the problems above, on the base of existing theory and algorithms, according to the characteristics of the volume of the passenger, the thesis presents an approach that predicts the volume of the passenger travel on the airplane based on genetic algorithm optimized Support Vector Machine. The thesis analyzes the basic ideas of Support Vector Machine and the basic principle of genetic algorithm. Improvements operating on the standard Genetic Algorithm What is more, it combines with the regression model of Support Vector Machine in case of nonlinear, on which to find out Support Vector Machine in the optimal kernel parameter using Genetic Algorithm, Finally, Support Vector Machine gives the forecasting effect analysis. Experimental content mainly includes:comparison with standard Genetic Algorithm and improve Genetic Algorithm comparison with Artificial Neural Networks Forecast prediction method as well as that of improved Support Vector Machine’prediction effects owing to different kernel function. The experimental results show the effectiveness and superiority of the method presented in the thesis.