The high–performance neural network system for monitoring of state and behavior of spacecraft subsystems by telemetry data

In work the neural network system of control and diagnostics of subsystems of space vehicles is described. The architecture, basic methods and principles of the software package operation are presented. The efficiency of artificial neural networks in the tasks of monitoring, diagnosing and predicting the behavior of subsystems of the spacecraft are shown. Experimental studies confirming the quality targets were carried out. (In Russian).