Currently, numerical simulations of seismic channel waves for the advance detection of geological structures in coal mine roadways focus mainly on modeling two- dimensional wave fields and therefore cannot accurately ...Currently, numerical simulations of seismic channel waves for the advance detection of geological structures in coal mine roadways focus mainly on modeling two- dimensional wave fields and therefore cannot accurately simulate three-dimensional （3-D） full-wave fields or seismic records in a full-space observation system. In this study, we use the first-order velocity-stress staggered-grid finite difference algorithm to simulate 3-D full-wave fields with P-wave sources in front of coal mine roadways. We determine the three components of velocity Vx, Vy, and Vz for the same node in 3-D staggered-grid finite difference models by calculating the average value of Vy, and Vz of the nodes around the same node. We ascertain the wave patterns and their propagation characteristics in both symmetrical and asymmetric coal mine roadway models. Our simulation results indicate that the Rayleigh channel wave is stronger than the Love channel wave in front of the roadway face. The reflected Rayleigh waves from the roadway face are concentrated in the coal seam, release less energy to the roof and floor, and propagate for a longer distance. There are surface waves and refraction head waves around the roadway. In the seismic records, the Rayleigh wave energy is stronger than that of the Love channel wave along coal walls of the roadway, and the interference of the head waves and surface waves with the Rayleigh channel wave is weaker than with the Love channel wave. It is thus difficult to identify the Love channel wave in the seismic records. Increasing the depth of the receivers in the coal walls can effectively weaken the interference of surface waves with the Rayleigh channel wave, but cannot weaken the interference of surface waves with the Love channel wave. Our research results also suggest that the Love channel wave, which is often used to detect geological structures in coal mine stopes, is not suitable for detecting geological structures in front of coal mine roadways. Instead, the Rayleigh channel wave can be used for the advance detection of geological structures in coal mine roadways.展开更多
This study combined fault identification with a deep learning algorithm and applied a convolutional neural network(CNN)design based on an improved balanced crossentropy(BCE)loss function to address the low accuracy in...This study combined fault identification with a deep learning algorithm and applied a convolutional neural network(CNN)design based on an improved balanced crossentropy(BCE)loss function to address the low accuracy in the intelligent identification of seismic faults and the slow training speed of convolutional neural networks caused by unbalanced training sample sets.The network structure and optimal hyperparameters were determined by extracting feature maps layer by layer and by analyzing the results of seismic feature extraction.The BCE loss function was used to add the parameter which is the ratio of nonfaults to the total sample sets,thereby changing the loss function to find the reference of the minimum weight parameter and adjusting the ratio of fault to nonfault data.The method overcame the unbalanced number of sample sets and improved the iteration speed.After a brief training,the accuracy could reach more than 95%,and gradient descent was evident.The proposed method was applied to fault identification in an oilfield area.The trained model can predict faults clearly,and the prediction results are basically consistent with an actual case,verifying the effectiveness and adaptability of the method.展开更多
基金supported by National Natural Science Foundation of China(Nos.41204077,41372290,41572244,51034003,51174210,and 51304126)natural science foundation of Shandong Province(Nos.ZR2011EEZ002 and ZR2013EEQ019)State Key Research Development Program of China(No.2016YFC0600708-3)
文摘Currently, numerical simulations of seismic channel waves for the advance detection of geological structures in coal mine roadways focus mainly on modeling two- dimensional wave fields and therefore cannot accurately simulate three-dimensional （3-D） full-wave fields or seismic records in a full-space observation system. In this study, we use the first-order velocity-stress staggered-grid finite difference algorithm to simulate 3-D full-wave fields with P-wave sources in front of coal mine roadways. We determine the three components of velocity Vx, Vy, and Vz for the same node in 3-D staggered-grid finite difference models by calculating the average value of Vy, and Vz of the nodes around the same node. We ascertain the wave patterns and their propagation characteristics in both symmetrical and asymmetric coal mine roadway models. Our simulation results indicate that the Rayleigh channel wave is stronger than the Love channel wave in front of the roadway face. The reflected Rayleigh waves from the roadway face are concentrated in the coal seam, release less energy to the roof and floor, and propagate for a longer distance. There are surface waves and refraction head waves around the roadway. In the seismic records, the Rayleigh wave energy is stronger than that of the Love channel wave along coal walls of the roadway, and the interference of the head waves and surface waves with the Rayleigh channel wave is weaker than with the Love channel wave. It is thus difficult to identify the Love channel wave in the seismic records. Increasing the depth of the receivers in the coal walls can effectively weaken the interference of surface waves with the Rayleigh channel wave, but cannot weaken the interference of surface waves with the Love channel wave. Our research results also suggest that the Love channel wave, which is often used to detect geological structures in coal mine stopes, is not suitable for detecting geological structures in front of coal mine roadways. Instead, the Rayleigh channel wave can be used for the advance detection of geological structures in coal mine roadways.
基金supported by the Natural Science Foundation of Shandong Province(ZR202103050722).
文摘This study combined fault identification with a deep learning algorithm and applied a convolutional neural network(CNN)design based on an improved balanced crossentropy(BCE)loss function to address the low accuracy in the intelligent identification of seismic faults and the slow training speed of convolutional neural networks caused by unbalanced training sample sets.The network structure and optimal hyperparameters were determined by extracting feature maps layer by layer and by analyzing the results of seismic feature extraction.The BCE loss function was used to add the parameter which is the ratio of nonfaults to the total sample sets,thereby changing the loss function to find the reference of the minimum weight parameter and adjusting the ratio of fault to nonfault data.The method overcame the unbalanced number of sample sets and improved the iteration speed.After a brief training,the accuracy could reach more than 95%,and gradient descent was evident.The proposed method was applied to fault identification in an oilfield area.The trained model can predict faults clearly,and the prediction results are basically consistent with an actual case,verifying the effectiveness and adaptability of the method.
