Yefremov S. Methods of split in the tasks of linear models of flight identification

The thesis is devoted to increasing the efficiency of a flying device movement mathematical model construction by the observed results on the basis of decomposition (splitting) realization. The work is based on the complex approach combining application of differential equations systems splitting and parametrical identification of objects mathematical model methods. The complexity of such a combination is that it is impossible to realize the splitting at unknown parameters of the flying device movement mathematical model. In this connection the splitting into subsystems of smaller dimensions equations is carried out on apriori estimations of an initial model's parameters. The date of flying tests are recalculated into split subspaces of the flying device movement components conditions, appropriate to subsystems of equations, to form the parameters of identification quality. The parameters of subsystems of equations are identified. The values of inexact known parameters of initial mathematical model offlight are specified by means of found subsystems' parameters estimations. The numerical way shows the convergence of iterative procedure's estimations of mathematical model parameters of flight.