Search

Coal burst prediction is an important research hotspot in coal mine production safety. This paper presents FDNet, which is a knowledge and data fusion-driven deep neural network for coal burst prediction. The main idea of FDNet is to extract explicit features based on the existing mine seismic physical model and utilize deep learning to automatically extract the implicit features of mine microseismic data. The key innovations of FDNet include an expert knowledge indicator selection method based on a subset search strategy, a mine microseismic data extraction method based on a deep convolutional neural network, and a feature deep fusion method of mine microseismic data based on an attention mechanism. We conducted a set of engineering experiments in Gaojiapu Coal Mine to evaluate the performance of FDNet. The results show that compared with the state-of-the-art data-driven machines and knowledge-driven methods, the prediction accuracy of FDNet is improved by 5% and 16%, respectively.

The development of the coal pore system is extremely significant for gas disaster prevention in coal mines and coalbed methane exploitation. To accurately, quantifiable, and non-destructively characterize the coal pore system, the relationship between total porosity, effective porosity, pore size distribution, and resistivity of coals with different degrees of metamorphic was studied by resistance test and low field NMR techniques. The dimensionless parameter ρP that characterize the coal pore system was defined by the variety of resistivity in dry and saturated conditions. The achieved results are presented as follows: (1) The greater the ρP (the influence of the coal pore system on the conductivity), the larger the total porosity; (2) The proportion of effective pores increased linearly with the addition of the ρP; (3) The proportion of micropores and small pores decreased linearly as the ρP increased, revealed that these pores had poor connectivity and complicated structure, and had a significant negative effect on the conductivity of saturated coal. (4) The proportion of mesopores and macropores gained linearly as the ρP increased, indicated these pores had well connectivity, and had a positive influence on the conductivity. The results showed that the variety of coal resistivity from dry to the saturated condition was mainly controlled by the pore system. Therefore, a new method to characterize the coal pore system was proposed, which was exceedingly effectively, conveniently, and non-destructively. [ABSTRACT FROM AUTHOR]