Research On Low Visibility Forecasting And Its Correlation With Flight Punctuality

Air transportion is one of the important modes of transportation in modern society.With the popularization of aviation,the safety and punctuality of flight have been generally concerned by the community.However,in southwest China,low visibility caused by fog or haze have became a significant factor affecting flight safety as well as punctuality.It appears that it is necessary forecasting the coming low visibility wearher accurately,therefore countermeasures would be taken before it comes to improve the punctuality during low visibility weather.Two main sides,low visibility forecasting and its correlation with flight punctuality,were abstracted as follows:For the low visibility forecast side,firstly the paper summarized the researches carried by scholars of various countries.Visbility is calssified into 3 grade according to aviation standards and be predicted by program based on BP Neural Network(BPNN)whose feasibility is confirmed through the analysis above.Secondly,statistics was made using meteorological data estracted from METAR reports of Chengdu Shuangliu International Airport during 2015~2017 in order to verify the distribute regulation of meteorological factors.Significance test was conduced then to determin the forecast index which would be used for BPNN construction.Finally,618 samples were extracted in random for BPNN model training,then the visibility grade in the future 3 houre could be predicted hourly through the model whose performance is in expectation.For the correlation between visibility and flight punctuality side.Due to the different of time scale selected for analysising,there are 2 chapters in this part: daily sacle analysis and hourly scale analysis.Firstly,serval indexes,such as flight punctual rate,which describe the normality of flight operation was set according to relative industry documents.Secondly discuss the correlation between these indexes and visibility in daily scale via statistics and scatter plotting.The result shows that visibility lower than 800 m can result in poor flight punctuality and its non-linear correlation is significantly.Finally,in hourly scale analysis,a concept called recovery time,which could assess the time spent for aerodrome air traffic flow recovering to normal at low visibility weather,was proposed.After example demonstration,the relationship between the recovery time and flight punctuality was obtained.The paper then give some suggestion for improving the flight punctuality performance and speeding the air traffic flow during low visibility weather.The result of the paper practically proves the applicability of low visibility forecasting based on BPNN.Also discover the correlation between visibility and flight punctuality index in quantitatively through data analysis.From the predicted visibility from model,situation of punctuality in next few hours could be estimated in advance.When the predictable low visibility weather upcoming in a few hours,stations as airlines,air traffic control units and aerodrome departments would make some pre-plans based on the predicted visibility grade and flight flow before it comes.The research is significative and pragmatic in ensuring the flight safety and improving the performance of flight punctuality in a certain degrees.