INFORMATION PROCESSING AND DATA ANALYSIS

Intelligent telemetry data analysis for diagnosing of the spacecraft hardware

N. S. Abramov, A. A. Talalaev, V. P. Fralenko

Ailamazyan Program Systems Institute of Russian Academy of Sciences

Abstract: This paper provide a method of spacecraft telemetry data mining, solves the problem of predicting and diagnosing its subsystems state. To solve these problems, the methods of pretreatment telemetry information using artificial neural networks of various configurations and modifications ZET-algorithm are offered. The results of clustering telemetry information by Kohonen neural network are described. It is shown that the developed methods are capable to solving tasks of monitoring and diagnostics of subsystems for spacecraft by telemetry data.

Keywords: spacecraft, monitoring, control, diagnostics, forecasting, method, neural networks, intelligent analysis system, telemetry data.