The Forecast Of Airport Peak Service Rate Based On Weather

In recent years,with the development of Chinese economy,the civil aviation transportation industry has also developed rapidly.However,its rapid development has also caused a series of problems and drawbacks.Nowadays,air traffic congestion is becoming more and more prominent,and there is large-scale flights delay.This not only makes the number of abnormal flights increase,but also leads to the decline of civil aviation safety and service quality.At the same time,the global climate has deteriorated sharply in recent years.Severe weather has become an important cause of the sharp decline in airport capacity,resulting in unbalanced flow and flight delays.In this context,more accurate and real-time prediction of airport traffic capacity under the influence of weather can rationalize air traffic flow,maximize the use of airspace resources and mitigate the impact of severe weather.Based on the previous studies,this paper proposes a relatively new method for predicting airport capacity.Firstly,the airport peak service rate and the airport peak service capacity reflecting the airport capacity are defined.Then,the correlation between the extracted weather data and the airport peak service rate is analyzed,and the weather data with high degree of influence is summarized.After the basic work is done,this paper first analyzes the airport peak service capability based on weather classification,uses k-means clustering algorithm to cluster the weather data,and matches the airport peak service rate,which is calculated.Airport peak service capacity corresponding to various weather conditions.Because the simple clustering effect on weather data is not good,this paper uses the random forest algorithm to obtain the weight of each weather element,and optimizes the clustering model,and finally obtains reasonable and ideal results.The airport peak service capacity under different weather types is used as a constraint condition,and it is added to the airport peak service rate prediction model considering the time dimension.In the prediction of airport peak service rate,this paper uses two machine learning algorithms,random forest algorithm and BP neural network algorithm,both of which predict the airport peak service rate.