Models and Technique Development for Spacecraft Control System Operational Reliability Analysis and Prognosis Based on Fuzzy Sets and Neural Networks

Control and navigation systems


Kosinskiy M. Y.*, Shatsky M. A.

Moscow Experimental Design Bureau “Mars”, 1-st Shemilovsky lane 16, building 2, Moscow, 127473, Russia



Spacecraft onboard control system (SOCS) is a complex multicomponent set of devices. It includes both hardware and software components.
Strict requirements applied to the reliability of such systems cause the necessity of creation and further development of quick methods for technical condition assessment during onboard control systems operation. SOCS operates under rough conditions of the outer space, so it is exposed to such disturbances as space radiation and wide range temperature changes.
The new approach to solving problem of SOCS reliability analysis based on combined application of fuzzy and neuro-fuzzy models is proposed in the paper.
The use of classical statistical approaches to the analysis of SOCS reliability at the operational phase appears to be inefficient. But, good results were obtained by using fuzzy logic to describe weakly formalizing relationships. Also, artificial neural networks could be applied to use data accumulated during operation for further reliability analysis.
The proposed hybrid model of reliability takes into account dispersion of characteristics of onboard control systems’ elements and influence of external factors. On the one hand it helps to compensate the lack of data for analysis during initial period of spacecraft operation and to perform more accurate analysis in the future using collected during operation data on the other hand.
The proposed approach is an alternative to the generally accepted at the present time methods of analysis of system reliability based mainly on statistical approach.
The technique was implemented as a PC software application and proved its effectiveness in SOCS reliability analysis during maintenance of existing spacecrafts.


reliability, fuzzy logic, neural networks, spacecraft, control system


  1. GOST 27.002-89 «Nadezhnost’ v technike. Osnovnye ponyatiya. Terminy i opredeleniya» (State Standard 27.002-89 «Reliability in Technics. General Concepts. Terms and Definitions»), Moscow, Standarty, 1990, 40 p.
  2. Brovkin А.G., Burdygov B.G., Gordiyko S.V. Bortovye sistemy upravleniya kosmicheskimi apparatami (Spacecraft Onboard Control Systems: Study Guide). Мoscow, MAI-PRINT, 2010, 304 p.
  3. Wang L.-X., Adaptive Fuzzy Systems and Control: Design and Stability Analysis, Prentice Hall, 1994, 232 p.
  4. Jang, J.-S. R. Structure Determination in Fuzzy Modeling: A fuzzy CART approach, Proceedings of International Conference on Fuzzy Systems, Orlando, Florida, June, 1994, pp. 480-485.

Download — informational site MAI

Copyright © 2000-2022 by MAI