A pattern recognition method for short-term forecast of the radiation situation in long-term manned space stations

This work is concerned with the possibilities of applying methods of pattern recognition for short-term forecast of the radiation situation during long-term stays in space stations in order to make reliable predictions of the radiation dose of the crew and its environment, which, presently, cannot be calculated with a sufficient accuracy. The method is based on the analysis of stochastic series from the monitoring data on the parameters of radiation conditions of the environment in modules of the Mir space station and the International Space Station (ISS).