Application Of Nonparametric Kernel Density Estimation In Fuel Policy Optimization

With the improvement of airlines' operational control capability and the capability of airports and ATC system.Reasonably reduce unnecessary fuel carrying on the premise of meeting the fuel policy,in particular,contingency fuel.It is conducive to promoting green and low-carbon flight,reducing fuel costs and enhancing the competitiveness of airlines.CCAR-121-R5 was approved by the ministry of transport,which had made changes to the fuel policy,and provided support to the regulations for the implementation o f airlines fuel policy optimization.Contingency fuel optimization is based on statistical methods.At the same time,how to accurately describe the characteristics of fuel consumption deviation data is a key step in the process of unanticipated fuel polic y optimization.So,to the data fitting part of the contingency fuel policy optimization.Firstly,a box chart anomaly testing model was established to detect,judge and eliminate abnormal data of fuel consumption deviation.Then,the integral mean square error method was used to obtain the optimal window width,and the fuel consumption deviation data was fitted by non-parametric kernel density estimation method,and the goodness of fit was tested by KS test.Finally,the fitted density curve was analyzed,and the corresponding quantile is determined according to the safety risk threshold,and the contingency fuel policy grade that the air carrier can apply to the CAAC is obtained.It provides the theoretical basis and reference for the application of the ai r carrier to implement the contingency fuel policy optimization.