Aimed at the characteristics of coal mine such as large quantity of information

much noises

many parameters and dynamic characteristics

etc.A coal mine environmental monitoring method of information fusion technology optimized by rough set data mining and D-S evidence theory are conducted.The rough set was carried out to deal with the information of coal mine.Distinguishing model of coal mine environment was established by using the radial basic function（RBF）neural network.Two-graded fusion decision was conducted by using D-S evidence theory to judge the security situation underground.The simulation results show that this method improved the identification and decision-making effects of coal mine environmental information

which greatly reduced the uncertainty.