A novel approach to coal and gas outburst prediction based on multi-sensor information fusion

Based on introduction of the background and the limitations of present prediction methods for gas outburst in coal mines, this paper focuses on introducing a new decision-making approach to coal and gas outburst prediction with multi-sensor information fusion. Two of the multi-sensor information fusion methods, neural network and the Dempster-Shafter evidence theory, were taken into account, and the improved combination rules of the D-S evidence theory in fuzzy sets was given for decision fusion. Then the practical experiment of gas outburst prediction is given to prove the efficiency and effectiveness of the new approach. The related experiments show that the novel approach with improved combination rules of the D-S evidence theory provides more rational results than each single prediction method.