Abstract: As the weathered bedrock and burned rock aquifers seriously threaten the production safety of mines in the Jurassic coalfield of northern Shaanxi Province, accurate prediction of their waterrichness and water inflow at the working face is of great significance for water control in mines. Aiming at the weathered bedrock and burned rock aquifers with close hydraulic connection, the area where working face 15217 of Shaanxi Hongliulin Coal Mine is located was taken as the study area. Meanwhile, the aquifer thickness, lithological combination index, index of rock burning and weathering degree and core recovery was taken as the evaluation indexes. The prediction method for the water richness of aquifers based on the support vector machine using manta ray foraging optimization was put forward. Then, the working face was zoned according to different waterrichness levels through the accurate zoned prediction of waterrichness of the weathered bedrock and burned rock aquifers. On this basis, the hydrogeological conditions of the working face before mining were analysed after a long period of underground dewatering, and the water inflow of different water-rich zones in the working face was predicted using the dynamic and static storage method and the error of water influx prediction results was small compared with that of the water influx measured in mining activities, ranging from 0.30 to 6.98 m³/h, which indicates that this prediction method has high feasibility and accuracy. It provides new ideas and methods for the prediction of water influx in the working faces of Hongliulin Coal Mine and mines with similar conditions. provides new ideas and methods for the prediction of water influx in the working faces of Hongliulin Coal Mine and mines with similar conditions.