Intellectual support of processes of control and diagnostics of space subsystems

We study the subject area is carried out, a review of existing developments in the field of constructing monitoring systems, monitoring and diagnostics of subsystems of spacecraft, including using the neural network approach. Theoretical studies was aimed at the implementation of mathematical and algorithmic support for the monitoring and diagnostics system of spacecraft subsystems. Our search for solutions was resulted in methodological approaches and methods for solving technical problems on the construction of a neural network monitoring system and diagnostics of subsystems of the spacecraft. The use of artificial neural network technologies makes it possible to detect, classify and predict errors, carry out multilevel diagnostics of subsystems of the spacecraft and predict their further behavior, thereby increasing the efficiency, speed of decision making and the reliability of the nodes of the spacecraft. The presented method of graphical representation of time sequences allows visual classification of the radio signal and noise detection. We propose to form and rank a set of significant features by applying the Add and Del algorithms.