Skip to Main Content
Flight Parameters stage classification is the premise of the fault diagnosis and trend forecast based on flight parameters. Stage classification belongs to the classification optimization problem of multi-attribute data through analysis the flight data. This paper carried out the research for the two-class classification based on the semi-supervised learning methods PTSVM (Progressive Transductive Support Vector Machines) and improved the PTSVM algorithm, which extends the application of PTSVM to the multi-class classification problem. The research and simulation work were carried out using the real flight parameters, and the comparison between the criterion of the flight parameters stage and the simulation results proved the validity of the research work for the flight parameters stage classification.