Research On Calculating Model Of Aircraft Runway Occupancy Time

With the recovery of China’s civil aviation industry,the operating pressure of airports has surged,especially the busy hub airports are facing greater demand for capacity.Improving runway capacity is the key to improve airport capacity.Studying and standardizing runway occupancy time is an important means to improve runway capacity.However,the existing research models on runway occupancy time are simple and the factors considered are not comprehensive,so it is of great significance and value to study and establish a new runway occupancy time calculation model.In this paper,based on the study of kernel principal component analysis,extreme learning machine and improved whale optimization algorithm,a computational model of runway occupancy time optimization of extreme learning machine using kernel principal component analysis to reduce dimension data based on improved whale algorithm was established to calculate accurate runway occupancy time.This model can provide reference for controllers to arrange flights and airports to construct new facilities.Firstly,the influence factors of runway occupation time and data sources were analyzed and introduced.Nine influence factors were selected from the Quick Access Recorder(QAR)data,and the QAR data were preprocessed and dimension reduced.To solve the problem that the element contribution rate is not high after the dimensionality reduction of principal component analysis,kernel principal component analysis is proposed to reduce the dimensionality of data.Secondly,based on extreme learning machine algorithm and BP neural network algorithm,two kinds of runway occupancy time calculation models are constructed.The model was tested and evaluated by using the QAR data of the test set,and the results showed that the calculation error of the extreme learning machine model was smaller.Therefore,the extreme Learning machine model was selected as the basic calculation model of this paper for further optimization.Then,after comparing and analyzing the advantages and disadvantages of the four intelligent optimization algorithms,the whale optimization algorithm is selected to improve.Aiming at the shortcomings of whale optimization algorithm,chaos theory,nonlinear convergence factor and crossbar strategy were used to improve the algorithm,and an improved whale optimization algorithm was established.The model of extreme learning Machine was optimized by using the improved whale optimization algorithm,and the computing model of the optimized extreme learning machine was established.Finally,this paper uses QAR data after dimensionality reduction of kernel principal component analysis as model input to establish a runway occupancy time calculation model based on improved whale algorithm to optimize extreme learning machine.The test set data was used to evaluate the model established in this paper,and the calculated value of the model and the runway occupancy time value recorded by QAR data were fitted,and the calculation errors of different models were quantified and compared.The experiment proved that the calculated results of the established runway occupancy time calculation model were closer to the actual value and the error was minimum,indicating that the model had a good effect.