Dispatch Release Intelligence Based On Particle Swarm Optimization Decision Research

In the daily operations of airlines,operational control and management play a crucial role,determining the operational safety and efficiency of airlines.With the continuous promotion of China’s "14th Five Year Plan" civil aviation development,the number of airlines is also increasing day by day,and their business scope is constantly expanding,and the requirements for the operational capacity of airlines are also continuously improving.In order to meet this demand,it is extremely important to improve the coordination efficiency and cooperation ability between various departments.Since the operation of civil aviation,the operation control centers at home and abroad have connected various departments with dispatcher as the hub,and people in the industry have been working hard to develop the dispatch automatic release system.So far,although Chinese airlines have their own control and operation system FOC,it has been more than a decade since its birth.Due to the limitations of algorithms at that time,although FOC can meet daily operations,it cannot further meet the rapidly growing number of flights and the service needs of the route network.Firstly,this article will construct a system of all factors involved in the dispatch and release process based on the currently running FOC system construction concept.During this process,all factors involved will be divided into three categories according to their data types,namely information that can be obtained in real-time,such as weather and altitude.The second category is information obtained from regulatory manuals,such as MEL and CDL manuals,The third type of data is calculated through formulas,and this type of data is mainly based on oil quantity.Secondly,after the establishment of the indicator system,in order to facilitate the construction of model constraints in the later stage,it is necessary to analyze all factors involved,and then quantify and homogenize all data based on a combination of objective and subjective methods.After completing this step,all data can be evaluated.This article uses the entropy weighted coefficient of variation method to evaluate all indicators during the research process.The purpose is to objectively and comprehensively reflect the role and importance of each influencing factor in the indicators.Finally,utilizing the excellent optimization decision-making ability of particle swarm optimization algorithm,a program capable of independently making intelligent dispatch and release decisions was developed.Because all factors were standardized in the research process of this article,a two-dimensional planar restricted area can be established after the establishment of the particle swarm algorithm model,allowing particles to optimize within the constrained range,and a release threshold can be obtained through thousands of iterations.This data processing process is faster and more accurate than previous algorithms.In order to improve release efficiency,make more reasonable dispatch release decisions,and connect various departments of the airline more efficiently,making more accurate judgments in the formulation of flight plans,analysis and handling of some emergency situations,and dynamic reception and analysis of aviation meteorology.At the same time,in order to further reduce the error rate during the iteration process,a deep residual network is used to optimize the algorithm,so that it can still maintain high accuracy after multiple iterations.