Accurate prediction of short-term rockburst has a significant role in improving the safety of workers in mining and geotechnical projects. The rockburst occurrence is nonlinearly correlated with its influencing factors that guarantee imprecise predicting results by employing the traditional methods. In this study, three approaches including including t-distributed stochastic neighbor embedding (t-SNE), K-means clustering, and extreme gradient boosting (XGBoost) were employed to predict the short-term rockburst risk. A total of 93 rockburst patterns with six influential features from micro seismic monitoring events of the Jinping-II hydropower project in China were used to create the database. The original data were randomly split into training and testing sets with a 70/30 splitting ratio. The prediction practice was followed in three steps. Firstly, a state-of-the-art data reduction mechanism t-SNE was employed to reduce the exaggeration of the rockburst database. Secondly, an unsupervised machine learning, i.e., K-means clustering, was adopted to categorize the t-SNE dataset into various clusters. Thirdly, a supervised gradient boosting machine learning method i.e., XGBoost was utilized to predict various levels of short-term rockburst database. The classification accuracy of XGBoost was checked using several performance indices. The results of the proposed model serve as a great benchmark for future short-term rockburst levels prediction with high accuracy.
23 April 2025 in International Journal of Coal Science & Technology
The fractured rock mass inherently exhibits uncertainty due to the presence of pre-existing discontinuities. In this study, a particle-based model incorporating the discrete fracture network (DFN) to elucidate the dynamic tensile responses and associated uncertainty of rock mass. At first, the particle-based model was used synthesize the intact rock and split Hopkinson pressure bar (SHPB) system, while the fractures were represented using the smooth fracture model (SJM). Subsequently, the samples of the fractured rock mass with varying joint geometrical configurations were conducted the dynamic tensile test using the numerical SHPB system. The simulated results demonstrate a gradual decrease in dynamic tensile strength (TS) with increasing fracture intensity and fracture length, which can be effectively described by nonlinear exponential functions. Additionally, the fracture orientation significantly influences the dynamic TS, however, the anisotropic characteristics gradually diminish as the deviation angle approaches 90°. Furthermore, as fracture intensity and fracture length increase, the dynamic TS variability also rises steadily. However, no noticeable pattern is seen when considering cases with varying fracture orientations. When subjected to SHPB loading, the fractured rock mass primarily exhibits a combined tensile-shear failure mode, contrasting with the pure tensile failure mode exhibited by the intact rock. These findings contribute significantly to comprehending the dynamic tensile responses of the fractured rock mass and can further enhance the stability analysis of in-situ rock engineering.
The traditional acoustic emission (AE) source location method, which is based on the model of a single-layered medium, is not appropriate for accurately locating AE sources in multilayered media. To solve this problem, an innovative AE source location method based on a refraction path method (RP) is proposed. In this new method, the refraction points at each interface are first solved according to Snell’s law and then substituted into the equations for the time difference of arrival (TDOA). Second, linear equations can be obtained by linearizing the TDOA equations. Finally, the optimal AE source can be obtained by iteratively updating the trial solution based on the linearized equations. The results of a pencil-lead break experiment and a numerical simulation show that the proposed method is applicable for locating the actual AE source in both single- and multilayered media, and the location error of this method in multilayered media is significantly smaller than that of traditional method.
In this paper, an improved automatic method for picking arrival times of acoustic emission (AE) signals, has been proposed to overcome the interference of spike noise and channel crosstalk. First, median filter is applied to eliminate spike noise presented in signals. Second, the preliminary arrival time is obtained by the AIC picker. Then a judgment whether the signal is disturbed by channel crosstalk is conducted. If the signal is triggered normally, the preliminary arrival time is regarded as the final arrival time of the AE signal. If the signal is triggered in advance, the signal series is selected from the preliminary arrival time to the time corresponding to the maximum of absolute values of signal series. At last the AIC picker is used again to pick the arrival time of selected series and the second AIC arrival time is interpreted as the final arrival time. To validate the performance of the improved automatic method, AIC picker, AIC picker applied after the median filter (MF-AIC) are also used simultaneously to pick the arrival times of AE signals and perform event location. The results show that the improved automatic method has better waveform adaptability and location accuracies than other two methods.