Abstract: To predict the water inrush risk of floor in deep coal seam mining, theoretical analysis, similar physical modeling and other research methods are used to conduct the intelligent identification of floor water inrush path. The prediction contour of floor water inrush is obtained based on the water pressure-stress monitoring data during the advance of mining face. According to the water pressure and stress data of each monitoring station, the probability index of water inrush at each position is quantitatively calculated. It is found that the greater this probility index, the higher the risk of water inrush. Taking the spatial distribution of water inrush probability index as the constraint condition, the A-Star path planning algorithm is used for the effective identification of the spatial path of water-conducting fracture, and an intelligent identification system of water inrush path is developed. Finally, similar material simulation was conducted for validation purpose. It is found that the path characteristics planned by the proposed method are highly consistent with the actual water-conducting fracture evolution trajectory of the model, and are highly consistent with the electrical monitoring results, verifying the effectiveness of the A-Star algorithm in water inrush path identification. The research results provide a new technical approach for real-time monitoring and early warning of floor water inrush disaster and guarantee safety of intelligent mine under deep mining conditions.