Abstract:Landslide susceptibility mapping is the first and most important step involved in landslide hazard assessment. The purpose of the present study is to compare three nonlinear approaches for landslide susceptibility mapping and test whether coal mining has a significant impact on landslide occurrence in coal mine areas. Landslide data collected by the Bureau of Land and Resources are represented by the X, Y coordinates of its central point; causative factors were calculated from topographic and geologic maps, as well as satellite imagery. The five-fold cross-validation method was adopted and the landslide/non-landslide datasets were randomly split into a ratio of 80:20. From this, five subsets for 20 times were acquired for training and validating models by GIS Geostatistical analysis methods, and all of the subsets were employed in a spatially balanced sample design. Three landslide models were built using support vector machine (SVM), logistic regression (LR), and artificial neural network (ANN) models by selecting the median of the performance measures. Then, the three fitted models were compared using the area under the receiver operating characteristics (ROC) curves (AUC) and the performance measures. The results show that the prediction accuracies are between 73.43% and 87.45% in the training stage, and 67.16% to 73.13% in the validating stage for the three models. AUCs vary from 0.807 to 0.906 and 0.753 to 0.944 in the two stages, respectively. Additionally, three landslide susceptibility maps were obtained by classifying the range of landslide probabilities into four classes representing low (0-0.02), medium (0.02-0.1), high (0.1-0.85), and very high (0.85-1) probabilities of landslides. For the distributions of landslide and area percentages under different susceptibility standards, the SVM model has more relative balance in the four classes compared to the LR and the ANN models. The result reveals that the SVM model possesses better prediction efficiency than the other two models. Furthermore, the five factors, including lithology, distance from the road, slope angle, elevation, and land-use types, are the most suitable conditioning factors for landslide susceptibility mapping in the study area. The mining disturbance factor has little contribution to all models, because the mining method in this area is underground mining, so the mining depth is too deep to affect the stability of the slopes.

No abstract

No abstract

The selective hydrogenation of C2H2 on five types of Cu(0)/Cu­(I), Cu(0), Cu­(I), PdCu(0), and PdCu­(I) catalysts are investigated to illuminate the effects of Cu component and the promoter on the selectivity and activity toward C2H4 formation. The selectivity and activity toward C2H4 formation on five types of catalysts are examined using density functional theory calculations. The results indicate that Cu(0)/Cu­(I) bicomponent catalyst presents higher selectivity but significantly lower activity toward C2H4 formation compared to the individual Cu(0) or Cu­(I) catalysts, suggesting that Cu(0)/Cu­(I) bicomponent catalyst cannot be selected as an ideal catalyst applied to the selective hydrogenation of C2H2, namely, the catalytic activity of Cu catalyst toward C2H4 formation dominates on the single-component rather than the multicomponent. On the other hand, for the promoter Pd-modified Cu(0) or Cu­(I) bimetallic catalysts, PdCu­(I) catalyst presents poor selectivity toward C2H4 formation compared to the individual Cu­(I) catalyst, whereas PdCu(0) catalyst exhibits better selectivity and activity toward C2H4 formation compared to the individual Cu(0) catalyst, indicating that the promoter-modified Cu catalysts should focus on the Cu(0) catalyst rather than the Cu­(I) catalyst. The results not only understand the reaction theories associated with the experimental results but also provide the basic theoretical clues for designing more efficient Cu-based catalysts applied in the selective hydrogenation of C2H2.

No abstract

No abstract

The Knothe time function is a classical method in predicting the ground mining subsidence. Nevertheless, it does not take the observation data into account in the prediction process. The Kalman filter method can solve this issue at large extent. Taking a coal mining work face of Xishan Coalfield as an example, this research compares the performance of the traditional Knothe time function and that of the improved Knothe time function by using the Kalman filter method. The comparison results show that through an improvement by using the Kalman filter method, the RMSE is improved from 133.4 mm to 78.3 mm; ME, from 91.9 mm to 3.1 mm; and the relative error, from 8.1% to 5.7%. Meanwhile, the improved model has good astringency. These verify that the improved model has higher accuracy and reliability. Hence, this research presents an effective method in predicting ground subsidence of mining area by improving the Knothe time function using the Kalman filter method.

Extreme learning machine (ELM), as an emergent technique for training feed-forward neural networks, has shown good performance on various learning domains. This work evaluates the effectiveness of a new Gaussian kernel function-based extreme learning machine (KELM) algorithm for the deformation prediction of mine slope surface utilizing various kinds of meteorological influence factor data including the temperature, atmospheric pressure, cumulative rainfall, relative humidity and refractive index of the mining slope. The KELM model was applied to the deformation of Anjialing diggings, which is an open-pit mine of the China Coal PingShuo Group Co., Ltd. in China. The prediction performance on real data suggests that the proposed KELM model can effectively express the non-linear relationship between the mine open-pit slope surface deformation and various kinds of meteorological influence factors. The prediction results are compared with the most widely used algorithms – Support vector machine (SVM) and back-propagation neural networks (BP NN) in terms of the ease of use ( for example, the number of user-defined parameters), regression accuracy and computation cost. The comparison shows that the new algorithm is similar to SVM and BP NN but more accurate, and has notable lower computational cost and stronger generalization ability.