Research And Implementation Of The Method For Flight Telegram’s Exception Handling

With fast development of air transportation, the flights in the air are soaring. Problems related to air traffic controlling become the key to the quality of air transportation. Automate management should be implemented. At present, in the air traffic management centers, AFTN is the main medium of information exchanging. But AFTN telegram is written by human whose standard is updating and different people who have different habit of writing that leads to low accuracy rate. Redundancy, losing, exception occur frequently that gives air traffic centers pressure.Aiming at the problem, firstly telegram format of civil air transportation which is published latest is analyzed. And summarize the exceptions during sending and receiving telegrams in real scenario. After researching the result of telegram exception, it is divided into two parts those are format exception and semantics exception. In processing the exceptions of formats, morphology is defined according to characteristic of telegram grouping and property is set on terminal symbol and non-terminal symbol using modified context-free grammar according to constitute of telegram, and grammar of telegram is defined. Repository of handling the exceptions is built, and repository method basing neural network is used to training data. Exceptions are settled by using text similarity matching method based on locating format exception. In processing the exceptions of semantics, model of semantics is defined. Finding the shortest path between two points and adding missing route information of telegram using A* algorithm basing on objective leading, and combine the relationship of semantic model to analyze semantics to finish processing exceptions.In this essay, research is based on format exceptions and semantics exceptions and a method of processing telegram information is come up. According to these two exceptions, the algorithm of handling the exceptions is implemented, and it is verified using telegram data collected from the northeast AENA. The result of experiment is counted and this algorithm is proved to be feasible and effective. This algorithm is flexible. If telegrams change it can modify the definition of morphology, grammar and semantics rules directly and program need not be updated. A* algorithm is used to complete the process of searching path when add information about route. The algorithm come up in this essay can solve the exceptions effectively, and has some contribution to handling telegram automatic.