Study on the Partial Paste Backfill Mining Method in a Fully Mechanized Top-Coal Caving Face: Case Study from a Coal Mine, China

Sustainability, Год журнала: 2024, Номер 16(11), С. 4393 - 4393

Опубликована: Май 22, 2024

Paste backfill mining is an significant part of green coal mining, which can improve resource utilization and extend the service life mines. It important for solving “three under, one above” problem avoiding industrial wastes such as gangue fly ash that occupy farmland pollute environment. To address difficult filling a fully mechanized top-coal caving face (FMT-CCF), new method partial paste herein proposed. First, implementation steps FMT-CCF are introduced in detail. Then, mechanistic model roof beam established. structural factors on effect 42105 determined. Dependent assay migration law overlying stratum after filling, numerical simulation analysis used to research feature main effect. Finally, rate, width, strength suitable obtained. When rate reaches 100%, alteration takes place, resulting efficient decrease rock stress arch shell’s height. As width body expands from 10 m at each end 20 m, experiences maximum reduction, specifically decreasing by approximately 14 m. greater than 0.4 GPa, better. This study has guidance reference significance thick seam mining.

Язык: Английский

---

Research on dynamic fuzzy prediction method for surrounding rock stability of mountain tunnels throughout the construction period

Tunnelling and Underground Space Technology, Год журнала: 2025, Номер 158, С. 106390 - 106390

Опубликована: Янв. 15, 2025

Язык: Английский

---

A coal bursting liability evaluation model based on fuzzy set theory and analysis of three influencing factors

Frontiers in Earth Science, Год журнала: 2024, Номер 12

Опубликована: Июнь 14, 2024

The classification of coal bursting liability is great significance for the prevention and control rock burst. To address shortcomings in existing methods, a comprehensive evaluation model based on combination weighted-fuzzy set theory three influencing factor analyses proposed. selects four indicators: dynamic failure time ( DT ), elastic energy index W ET K E uniaxial compressive strength R C ). Two types membership functions, trapezoidal fuzzy numbers (TFN) Gaussian (GFN), are used to quantitatively describe fuzziness between indicator levels. Delphi method random forest feature identification combined obtain subjective objective weighting, determining optimal weight indicators. Based Zadeh operator (ZO), maximum-minimum (MMO), weighted-average (WAO), all-around restrictive (ARO), calculations carried out synthesis weights memberships. Maximal principle (MMP) Credible (CIP) utilized as assess level, constructing 16 models. impact operators, results systematically analyzed discrimination 127 sample sets. show that constructed using numbers, weighted average operator, maximal (TFN-WAO-MMP), with accuracy 97.64%. Finally, applied 10 engineering instances, consistent actual situation, verifying reliability effectiveness model. Overall, these findings contribute development more sophisticated accurate assessing burst tendency specimens. By leveraging sets, this approach provides nuanced tendency, thus offers potential improve workplace safety efficiency mining industry.

Язык: Английский

---

Classification of coal bursting liability of some chinese coals using machine learning methods

Scientific Reports, Год журнала: 2024, Номер 14(1)

Опубликована: Июнь 18, 2024

The classification of coal bursting liability (CBL) is essential for the mitigation and management bursts in mining operations. This study establishes an index system CBL classification, incorporating dynamic fracture duration (DT), elastic strain energy (WET), (KE), uniaxial compressive strength (RC). Utilizing a dataset comprising 127 measurement groups, impacts various optimization algorithms were assessed, two prominent machine learning techniques, namely back propagation neural network (BPNN) support vector (SVM), employed to develop twelve distinct models. models' efficacy was evaluated based on accuracy, F1-score, Kappa coefficient, sensitivity analysis. Among these, Levenberg-Marquardt (LM-BPNN) model identified as superior, achieving accuracy 96.85%, F1-score 0.9113, coefficient 0.9417. Further validation Wudong Coal Mine Yvwu Industry confirmed model, 100% accuracy. These findings underscore LM-BPNN model's potential viable tool enhancing burst prevention strategies sectors.

Язык: Английский

---

Classification forecasting research of rock burst intensity based on the BO-XGBoost-Cloud model

Earth Science Informatics, Год журнала: 2024, Номер 18(1)

Опубликована: Дек. 21, 2024

Язык: Английский

---