Research On The Prediction Method Of Probabilistic Traffic Demand Of En-route Sector

With the increasing pressure brought by the increasing demand for air traffic,the airspace resources available to the civil aviation industry in China have become increasingly tense.The operational conflicts in the actual operation of the airspace have intensified,and airspace congestion has occurred frequently.Based on the advanced research results of foreign countries,combined with the practical problems of China's airspace congestion management,this paper uses air traffic management,air traffic flow management,traffic statistics and forecasting,nuclear density estimation,non-parametric nuclear density estimation and other theoretical methods,around the route The sector's traffic demand forecasting problem analyzes the main factors affecting sectoral traffic demand,establishes traffic demand forecasting models and methods from the angle of uncertainty,and carries out corresponding simulation verification.The main research content includes:Firstly,the concept of probabilistic traffic demand for the en-route sector was studied.The kernel density estimation theory and the non-parametric kernel density estimation method were introduced.The main reasons for applying the non-parametric kernel density estimation theory and the non-parametric kernel density estimation method were clarified.Secondly,from the aspects of sector structure,weather influence,time distribution of aircraft access sector,etc.,the disturbance factors affecting the traffic demand of the en-route sectors were analyzed,the main influence factors were extracted,combined with non-parametric kernel density estimation theory and non-parametric kernel density estimation.The method proposes a probabilistic traffic demand forecasting model for route sectors.Then,the data smoothing process,kernel density estimation function selection,bandwidth selection,and prediction error analysis were introduced.Probability traffic demand forecasting methods for route sector were proposed.The third is based on the proposed model and method for the probabilistic traffic demand forecasting of the route sector.Based on the real historical data of the route sector in the South Central region,the simulation verification and analysis of the model method was carried out,and the sector's future traffic was obtained.The demand value and its probability distribution,and found that the probabilistic traffic demand forecast result is more accurate than the traditional deterministic traffic demand forecasting method,and is suitable for providing demand forecasting data for high-altitude route congestion control strategies in China.