Optimization of Height Prediction of Water Fractured Zone Based on ANN
This paper discussed the factors that affect the height of water fractured zone, and they were divided into primary and secondary factors, in order to construct a system that included factors that affect the height of water fractured zone. By using the BP neural network model, this paper chose the thickness of coal seam, roof lithology, tilt angle of coal seam, overburden hardness, working face length, advance speed and rock bulking to be the primary factors in order to simplify the model and accelerate speed. If the mining geological condition was clear, we could ignore the secondary factors. Prediction results showed that the simplified BP model could meet the accuracy of the height prediction of water fractured zone and the prediction method could provide technical guidance and a certain safety for coal mining under water.