Research On The B737 Flight Fault Analysis And Decision-making Algorithm

With the continuous development of civil aviation, more and more Airline Companies and passengers increasingly become concerned for aircraft flight safety issues. This paper based on the B737 flight fault analysis and decision-making application background,develops the system of association rules mining based on QAR(Quick Access Recorder) data combining the aircraft maintenance manual, Research and implementation of algorithms for association rule mining and decision-tree is carried out. The main research contents are summarized as follows.1. The QAR data association rules mining system has been developed, which mainly consists of two major components, the data preprocessing module and the association rule mining module. Implicit relationships between some of the parameters of the B737 aircraft engine QAR Data species found by mining and this study provide practical value for QAR data applications.2. In order to mine the failure modes of knowledge hidden behind the QAR Data, after QAR data preprocessing, then draw on association rules algorithm thinking, an improved association rule algorithm based on the list of transactions is put forward, By scanning the QAR transaction database, Affairs list is generated, then layer iterative method through the low-dimensional frequent item sets high-dimensional frequent item sets, effective mining association rules between the QAR data parameters.3. For lack of knowledge acquisition in traditional decision system, By using the original fault diagnosis decision table, the diagnostic decision tree which constructed by decision tree algorithm transformed to the decision-making rule, which aided to flight fault diagnosis,enhance the reasoning of the fault diagnosis system diagnostic capabilities and the quality of diagnosis and meet the diagnostic needs better.