Rockburst is a common dynamic geological hazard, severely restricting the development and utilization of underground space and resources. As the depth of excavation and mining increases, rockburst tends to occur frequently. Hence, it is necessary to carry out a study on rockburst prediction. Due to the nonlinear relationship between rockburst and its influencing factors, artificial intelligence was introduced. However, the collected data were typically imbalanced. Single algorithms trained by such data have low recognition for minority classes. In order to handle the problem, this paper employed stacking technique of ensemble learning to establish rockburst prediction models. In total, 246 sets of data were collected. In the preprocessing stage, three data mining techniques including principal component analysis, local outlier factor and expectation maximization algorithm were used for dimension reduction, outlier detection and outlier substitution, respectively. Then, the pre-processed data were split into a training set (75%) and a test set (25%) with stratified sampling. Based on the four classical single intelligent algorithms, namely k-nearest neighbors (KNN), support vector machine (SVM), deep neural network (DNN) and recurrent neural network (RNN), four ensemble models (KNN–RNN, SVM–RNN, DNN–RNN and KNN–SVM–DNN–RNN) were built by stacking technique of ensemble learning. The prediction performance of eight models was evaluated, and the differences between single models and ensemble models were analyzed. Additionally, a sensitivity analysis was conducted, revealing the importance of input variables on the models. Finally, the impact of class imbalance on the prediction accuracy and fitting effect of models was quantitatively discussed. The results showed that stacking technique of ensemble learning provides a new and promising way for rockburst prediction, which exhibits unique advantages especially when using imbalanced data.
There is a well-known seesaw pattern of precipitation between the tropical western North Pacific(WNP) and the Yangtze River basin(YRB) during summer. This study identified that this out-of-phase relationship experiences a subseasonal change;that is, the relationship is strong during early summer but much weaker during mid-summer. We investigated the large-scale circulation anomalies responsible for the YRB rainfall anomalies on the subseasonal timescale. It was found that the YRB rainfall is mainly affected by the tropical circulation anomalies during early summer, i.e., the anticyclonic or cyclonic anomaly over the subtropical WNP associated with the precipitation anomalies over the tropical WNP. During mid-summer, the YRB rainfall is mainly affected by the extratropical circulation anomalies in both the lower and upper troposphere. In the lower troposphere, the northeasterly anomaly north of the YRB favors heavier rainfall over the YRB by intensifying the meridional gradient of the equivalent potential temperature over the YRB. In the upper troposphere, the meridional displacement of the Asian westerly jet and the zonally oriented teleconnection pattern along the jet also affect the YRB rainfall. The subseasonal change in the WNP–YRB precipitation relationship illustrated by this study has important implications for the subseasonalto-seasonal forecasting of the YRB rainfall. 相似文献
Two large earthquakes (an earthquake doublet) occurred in south-central Turkey on February 6, 2023, causing massive damages and casualties. The magnitudes and the relative sizes of the two mainshocks are essential information for scientific research and public awareness. There are obvious discrepancies among the results that have been reported so far, which may be revised and updated later. Here we applied a novel and reliable long-period coda moment magnitude method to the two large earthquakes. The moment magnitudes (with one standard error) are 7.95±0.013 and 7.86±0.012, respectively, which are larger than all the previous reports. The first mainshock, which matches the largest recorded earthquakes in the Turkish history, is slightly larger than the second one by 0.11±0.035 in magnitude or by 0.04 to 0.18 at 95% confidence level. 相似文献
基于实验室测量数据,构建了盐度和温度多项式形式的L波段海水介电模型,并采用最小二乘法和奇异值分解技术求解模型系数。将新模型与目前广泛使用的Klein-Swift(KS)模型、Meissner-Wentz(MW)模型以及George Washington University(GWU)模型进行比较。结果表明:新模型计算的海水介电常数与实验室测量数据的RMSE误差为0.09(实部)和0.25(虚部),均优于KS、MW、GWU模型。 相似文献
In this study, two series of physical modeling experiments, with and without a grouting process, were conducted under different grouting pressures to study the effect of compaction grouting on the performance of compaction-grouted soil nails. In addition, a hyperbola-based model was proposed to describe the variation of the pullout forces with and without grouting. Some of the main conclusions drawn are as follows. First, the compaction effect initially influences the mobilized pullout force, but not the final stage of pullout; the large difference between the two series of tests in regard to the pullout force at the initial stage led to the first part of this conclusion. However, the final pullout force results of the tests, both those with and those without grouting, were similar. Second, once the soil condition changes, the compaction effect on the performance of a soil nail depends on the grouting pressure rather than the diameter of the grout bulb. Third, the difference in the soil response (i.e., vertical dilatancy and the vertical and horizontal squeezing effects) derived from the compaction grouting effect will result in the initial difference in the increased rate of the pullout force between the tests with a grouting process and those without. Finally, a hyperbola-based model was proposed to describe the variation of the pullout force of the model tests with and without grouting, through which the pullout force is available of prediction for the given diameter of grout bulb and pullout displacement.