Research On Hazard Source Identification And Its Application In Civil Aviation Safety Control

The rapid growth of the social economy has promoted the rapid development of the civil aviation industry and the amount of traffic is increasing.The aircraft plays an important role in today's transportation.The issue that is inseparable from the airplane is the problem of security.Safety management is an important subject in civil aviation air traffic control.What's more,the accurate identification and effective analysis of hazards is an important guarantee and premise for security.This paper is based on the safety management system of civil aviation air traffic control,aiming at the hazard identification and analysis which is the key link of civil aviation safety management process,adopted improved online sequential learning machine and ant colony algorithm to mining association rules for identification and analysis,and designed the overall framework of the civil aviation safety management system according to actual demands,and the realization of the important function and key technologies in the system are given.The main work of this paper is as follows:The rapid growth of the social economy has promoted the rapid development of the civil aviation industry and the amount of traffic is increasing.The aircraft plays an important role in today's transportation.The main task of air traffic management is safety management.1.The requirements of civil aviation safety management system is analyzed,the overall framework of the system is designed,the key technology of the system is expounded,and the process of the main function module in the system is given.2.There are many environment influencing factors and various kinds of equipment with complex parameters during flight,and when crossing different geographic areas,the related data may change because of the change of the latitude,the longitude,and the weather,so the characteristic of hazard are numerous and hazard need to be judged in real time.Aiming at this problem,a hazard identification algorithm based on improved online sequential extreme learning machine is proposed.It utilizes hidden layer feed forward neural network to train hazard data,adds newly arrived data into the training network,and then updates the network outputs weight in real time,in order to realize the online identification.In addition,combining with adaptive differential evolution algorithm,using its good global optimization ability to optimize network input weights and hidden layer bias,it optimize the network structure to obtain more accurate recognition results.3.Aiming at the problem of much human participation and the lack of objectivity in the hazard causes analysis,a hazard cause analysis algorithm based on mixed ant colony association rules mining is proposed.It exploits association rule mining to explore the cause of hazard,and discards the shortcomings that the traditional association rule mining algorithms repeatedly scan the database leading to low efficiency and generate a large number of candidate sets,uses the improved ant colony algorithm to mine the maximal frequent item sets instead.Besides,the initial pheromone concentration of the algorithm is determined by particle swarm,in order to avoid the initial blindness of the single ant colony algorithm.4.The application background of civil aviation safety management system is introduced,based on the key technologies of civil aviation safety management system,the realization of the system functions such as hazard identification,reason analysis,alarm,inspection is given,and they have been initially applied into the civil aviation safety management system and achieved good results.