Evaluation Of Large Airport Cooperative Operation Based On Improved Support Vector Machine

The rapid development of modern science and technology has created an increasingly large airport transportation system,increasingly strict and standardized service level requirements,which also makes the evaluation of airport operation status more and more important.However,the evaluation of airport operation status is a complex problem because it involves many types of transport capacity,large scale of passenger flow and large amount of operation data.The operation state of airport hub is an important embodiment of the collection and distribution capacity of urban transportation hub,and plays an irreplaceable role in ensuring passenger transportation service.The normal operation of airport hub also directly affects the operation of the city.Therefore,the use of intelligent technology for state analysis and evaluation has become an important measure to ensure the efficient evacuation of passenger flow.To solve this problem,the main work of this paper is done as follows.First,the efficiency of airport coordinated operation is selected as the object.Based on the actual construction needs,exemplary needs,analysis previous researches and the actual needs of this research,several evaluation indexes are selected to evaluate the large airport transportation system,such as the transportation service index,infrastructure service index,exemplary index,sustainable development index and other additional index.After that,these indexes are divided into different levels and the selection method of evaluation indexes is proposed.Then,based on the correlation analysis of passenger satisfaction with the service quality of the airport cooperative operation system,the selected evaluation indexes are analyzed with the assistance of the Lasso regression.This method can simplify the calculation of the regression model,reduce the multicollinearity of explanatory variables and realize the efficient and accurate identification of the key indexes which significantly affects the system.Three important factors are obtained,including capacity coordinated matching degree,hub operation demonstration and average transfer time.Next,a support vector machine decision tree model is proposed to construct a three-level SVM classifier,and hence to divide the existing operation status data into test set and training set.The grid search and 50% cross validation methods are used to optimize the model parameters.Based on these measures,the evaluation method of large airport coordinated operation state is proposed to classify and evaluate the airport transportation system.Finally,based on the historical operating status data and real-time basic data set provided by Beijing Capital International Airport,the characteristics of airport operation and evaluation index system are analyzed,and the rationality and feasibility of the proposed large airport coordinated operating state evaluation method are verified.