Predicting Rock Burst in Underground Engineering Leveraging a Novel Metaheuristic-Based LightGBM Model

Rock bursts represent a formidable challenge in underground engineering,posing substantial risks to both infrastructure and human safety.These sudden and violent failures of rock masses are characterized by the rapid release of accumulated stress within the rock,leading to severe seismic events and structural damage.Therefore,the development of reliable prediction models for rock bursts is paramount to mitigating these hazards.This study aims to propose a tree-based model—a Light Gradient Boosting Machine(LightGBM)—to predict the intensity of rock bursts in underground engineering.322 actual rock burst cases are collected to constitute an exhaustive rock burst dataset,which serves to train the LightGBMmodel.Two population-basedmetaheuristic algorithms are used to optimize the hyperparameters of the LightGBM model.Finally,the sensitivity analysis is used to identify the predominant factors that may incur the occurrence of rock bursts.The results show that the population-based metaheuristic algorithms have a good ability to search out the optimal hyperparameters of the LightGBM model.The developed LightGBM model yields promising performance in predicting the intensity of rock bursts,with which accuracy on training and testing sets are 0.972 and 0.944,respectively.The sensitivity analysis discloses that the risk of occurring rock burst is significantly sensitive to three factors:uniaxial compressive strength(σc),stress concentration factor(SCF),and elastic strain energy index(Wet).Moreover,this study clarifies the particular impact of these three factors on the intensity of rock bursts through the partial dependence plot.

Prediction of rock burst classification using cloud model with entropy weight

The method of cloud model with entropy weight was adopted for the prediction of rock burst classification. Some main factors of rock burst including the uniaxial compressive strength （σc）, the tensile strength （σt）, the tangential stress （σθ）, the rock brittleness coefficient （σc/σt）, the stress coefficient （σθ /σc） and the elastic energy index （Wet） are chosen to establish evaluation index system. The entropy？cloud model and criterion are obtained through 209 sets of rock burst samples from underground rock projects. The sensitivity of indicators is analyzed and 209 sets of rock burst samples are discriminated by this model. The discriminant results of the entropy-cloud model are compared with those of Bayes, KNN and RF methods. The results show that the sensitivity order of those factors from high to low is σ_θ /σ_c, σ_θ, W_（ct）, σ_c/σ_t, σ_t, σ_c, and the entropy-cloud model has higher accuracy than Bayes, K-Nearest Neighbor algorithm （KNN） and Random Forest （RF） methods.

Prediction and control of rock burst of coal seam contacting gas in deep mining

By analyzing the characteristics and the production mechanism of rock burstthat goes with abnormal gas emission in deep coal seams,the essential method of eliminatingabnormal gas emission by eliminating the occurrence of rock burst or depressingthe magnitude of rock burst was considered.The No.237 working face was selected asthe typical working face contacting gas in deep mining;aimed at this working face,a systemof rock burst prediction and control for coal seam contacting gas in deep mining wasestablished.This system includes three parts:① regional prediction of rock burst hazardbefore mining,② local prediction of rock burst hazard during mining,and ③ rock burstcontrol.

Risk evaluation of rock burst through theory of static and dynamic stresses superposition

Rock burst is one of the most catastrophic dynamic hazards in coal mining. A static and dynamic stresses superposition-based(SDSS-based) risk evaluation method of rock burst was proposed to pre-evaluate rock burst risk. Theoretical basis of this method is the stress criterion incurring rock burst and rock burst risk is evaluated according to the closeness degree of the total stress(due to the superposition of static stress in the coal and dynamic stress induced by tremors) with the critical stress. In addition, risk evaluation criterion of rock burst was established by defining the "Satisfaction Degree" of static stress. Furthermore,the method was used to pre-evaluate rock burst risk degree and prejudge endangered area of an insular longwall face in Nanshan Coal Mine in China. Results show that rock burst risk is moderate at advance extent of 97 m, strong at advance extent of 97-131 m,and extremely strong(i.e. inevitable to occur) when advance extent exceeds 131 m(mining is prohibited in this case). The section of two gateways whose floor abuts 15-3 coal seam is a susceptible area prone to rock burst. Evaluation results were further compared with rock bursts and tremors detected by microseismic monitoring. Comparison results indicate that evaluation results are consistent with microseismic monitoring, which proves the method's feasibility.

Application of high-pressure water jet technology and the theory of rock burst control in roadway

This paper puts forward using high-pressure water jet technology to control rock burst in roadway, and analyzes the theory of controlling rock burst in roadway by the weak structure zone model. The weak structure zone is formed by using high-pressure water jet to cut the coal wall in a continuous and rotational way. In order to study the influence law of weak structure zone in surrounding rock, this paper numerically analyzed the influence law of weak structure zone, and the disturbance law of coal wall and floor under dynamic and static combined load. The results show that when the distance between high-pressure water jet drillings is 3 m and the diameter of drilling is 300 mm, continuous stress superposition zone can be formed. The weak structure zone can transfer and reduce the concentrated static load in surrounding rock, and then form distressed zone. The longer the high-pressure water jet drilling is, the larger the distressed zone is. The stress change and displacement change of non-distressed zone in coal wall and floor are significantly greater than that of distressed zone under dynamic and static combined load. And it shows that the distressed zone can effectively control rock burst in roadway under dynamic and static combined load. High-pressure water jet technology was applied in the haulage gate of 250203 working face in Yanbei Coal Mine, and had gained good effect. The study conclusions provide theoretical foundation and a new guidance for controlling rock burst in roadway.

Rock burst prediction based on genetic algorithms and extreme learning machine

Rock burst is a kind of geological disaster in rock excavation of high stress areas.To evaluate intensity of rock burst,the maximum shear stress,uniaxial compressive strength,uniaxial tensile strength and rock elastic energy index were selected as input factors,and burst pit depth as output factor.The rock burst prediction model was proposed according to the genetic algorithms and extreme learning machine.The effect of structural surface was taken into consideration.Based on the engineering examples of tunnels,the observed and collected data were divided into the training set,validation set and prediction set.The training set and validation set were used to train and optimize the model.Parameter optimization results are presented.The hidden layer node was450,and the fitness of the predictions was 0.0197 under the optimal combination of the input weight and offset vector.Then,the optimized model is tested with the prediction set.Results show that the proposed model is effective.The maximum relative error is4.71%,and the average relative error is 3.20%,which proves that the model has practical value in the relative engineering.

Prevention and forecasting of rock burst hazards in coal mines

Rock bursts signify extreme behavior in coal mine strata and severely threaten the safety of the lives of miners, as well as the effectiveness and productivity of miners. In our study, an elastic-plastic-brittle model for the deformation and failure of coal/rock was established through theoretical analyses, laboratory experiments and field testing, simulation and other means, which perfectly predict sudden and delayed rock bursts. Based on electromagnetic emission (EME), acoustic emission (AE) and microseism (MS) effects in the process from deformation until impact rupture of coal-rock combination samples, a multi-parameter identification of premonitory technology was formed, largely depending on these three forms of emission. Thus a system of classification for forecasting rock bursts in space and time was established. We have presented the intensity weakening theory for rock bursts and a strong-soft-strong (3S) structural model for controlling the impact on rock surrounding roadways, with the objective of laying a theoretical foundation and establishing references for parameters for the weakening control of rock bursts. For the purpose of prevention, key technical parameters of directional hydraulic fracturing are revealed. Based on these results, as well as those from deep-hole controlled blasting in coal seams and rock, integrated control techniques were established and anti-impact hydraulic props, suitable for roadways subject to hazards from rockbursts have also been developed. These technologies have been widely used in most coal mines in China, subject to these hazards and have achieved remarkable economic and social benefits.

Rock burst proneness prediction by acoustic emission test during rock deformation

Rock burst is a severe disaster in mining and underground engineering,and it is important to predict the rock burst risk for minimizing the loss during the constructing process.The rock burst proneness was connected with the acoustic emission(AE) parameter in this work,which contributes to predicting the rock burst risk using AE technique.Primarily,a rock burst proneness index is proposed,and it just depends on the heterogeneous degree of rock material.Then,the quantificational formula between the value of rock burst proneness index and the accumulative AE counts in rock sample under uniaxial compression with axial strain increases is developed.Finally,three kinds of rock samples,i.e.,granite,limestone and sandstone are tested about variation of the accumulative AE counts under uniaxial compression,and the test data are fitted well with the theoretic formula.

Rock burst induced by roof breakage and its prevention

Based on the research on rock burst phenomenon induced by the breakage of thick and hard roof around roadways and working faces in coal mines, a criterion of rock burst induced by roof breakage （RBRB） was proposed and the model was built. Through the model, a method calculating the varied stresses induced by roof breakage in support objects and coal body was proposed and a unified formula was derived for the calculation of stress increment on support objects and coal body under different breaking forms of roof. Whilst the formula for calculating dynamic load was derived by introducing dynamic index Kd. The formula was verified in Huating Mine by stress measurement. According to the formula for stress increment calculating, the sensitivities of dynamic load parameters were further studied. The results show that the thickness and breaking depth of roof, width of support objeet are the sensitive factors. Based on the discussion of the model, six associated effective methods for rock burst prevention are obtained.

Numerical Simulation of Rock Burst in Circular Tunnels Under Unloading Conditions

Rock burst in a circular tunnel under high in-situ stress conditions was investigated with a numerical method coupled the rock failure process theory (RFPA) and discontinuous deformation theory (DDA). Some numerical tests were carraied out to investigate the failuer patterns of circular tunnel under unloading conditions. Compared the results under loading conditions,the shapes of failure zones are more regular under the unloading conditions. The failure pat-terns in the same type of rock mass are clearly different because of non-homogeneity of the rock material. The extension of cracks shows some predictability with an increasing of in-situ stress. When the homogeneity index of rocks (m) is ei-ther relatively high or low and lateral pressure coefficients (λ) is high,the number of regular shear slide cracks decreases and the probability of a rock burst also becomes lower. Our numerical simulation results show that the stability of sur-face rock and the natural bedding stratification of rock material greatly affect rock bursts. Installing bolts with due dili-gence and suitably can effectively prevent rock bursts. However,it is not effective to control rock bursts by releasing the strain energy with normal pre-boreholes.

Numerical simulation study on hard-thick roof inducing rock burst in coal mine

In order to reveal the dynamic process of hard-thick roof inducing rock burst, one of the most common and strongest dynamic disasters in coal mine, the numerical simulation is conducted to study the dynamic loading effect of roof vibration on roadway surrounding rocks as well as the impact on stability. The results show that, on one hand, hard-thick roof will result in high stress concentration on mining surrounding rocks; on the other hand, the breaking of hard-thick roof will lead to mining seismicity, causing dynamic loading effect on coal and rock mass. High stress concentration and dynamic loading combination reaches to the mechanical conditions for the occurrence of rock burst, which will induce rock burst. The mining induced seismic events occurring in the roof breaking act on the mining surrounding rocks in the form of stress wave. The stress wave then has a reflection on the free surface of roadway and the tensile stress will be generated around the free surface. Horizontal vibration of roadway surrounding particles will cause instant changes of horizontal stress of roadway surrounding rocks; the horizontal displacement is directly related to the horizontal stress but is not significantly correlated with the vertical stress; the increase of horizontal stress of roadway near surface surrounding rocks and the release of elastic deformation energy of deep surrounding coal and rock mass are immanent causes that lead to the impact instability of roadway surrounding rocks. The most significant measures for rock burst prevention are controlling of horizontal stress and vibration strength.Key words

Determination of ejection velocity of rock fragments during rock burst in consideration of damage

The ejection velocity of rock fragments during rock burst, one of the important indexes representing the rock burst strength, is used most conveniently in the supporting design of tunnel with rock burst tendency and is often determined by means of observation devices. In order to calculate the average ejection velocity of rock fragments theoretically, the energy of rock burst was divided into damage consuming energy and kinetic energy gained by unit volume of rock firstly, and then the rock burst kinetic proportional coefficient η was brought up which could be determined according to the rock-burst damage energy index W_D , at last the expression of the average ejection velocity of rock fragments during rock burst was obtained and one deep level underground tunnel was researched using the mentioned method. The results show that the calculation method is valid with or without considering the tectonic stress of tunnels, and that the method can be a reference for supporting design of deep mining.

Quantitative calculation for the dissipated energy of fault rock burst based on gradient-dependent plasticity

The capacity of energy absorption by fault bands after rock burst wascalculated quantitatively according to shear stress-shear deformation curves considering theinteractions and interplaying among microstructures due to the heterogeneity of strain softeningrock materials. The post-peak stiffness of rock specimens subjected to direct shear was derivedstrictly based on gradient-dependent plasticity, which can not be obtained from the classicalelastoplastic theory. Analytical solutions for the dissipated energy of rock burst were proposedwhether the slope of the post-peak shear stress-shear deformation curve is positive or not. Theanalytical solutions show that shear stress level, confining pressure, shear strength, brittleness,strain rate and heterogeneity of rock materials have important influence on the dissipated energy.The larger value of the dissipated energy means that the capacity of energy dissipation in the formof shear bands is superior and a lower magnitude of rock burst is expected under the condition ofthe same work done by external shear force. The possibility of rock burst is reduced for a lowersoftening modulus or a larger thickness of shear bands.

Study on fault induced rock bursts

In order to study the rules of rock bursts caused by faults by means of mechanical analysis of a roof rock-mass balanced structure and numerical simulation about fault slip destabilization, the effect of coal mining operation on fault plane stresses and slip displacement were studied. The results indicate that the slip displacement sharply increases due to the decrease of normal stress and the increase of shear stress at the fault plane when the working face advances from the footwall to the fault itself, which may induce a fault rock burst. However, this slip displacement will be very small due to the increase of normal stress and the decrease of shear stress when the working face advances from the hanging wall to the fault itself, which results in a very small risk of a fault rock burst.

Gradient principle of horizontal stress inducing rock burst in coal mine

Based on the stress distribution characteristics of rock burst multiple sites, the criterion of horizontal stress inducing layer dislocation rock burst was established. Accordingly, the influencing factors were analyzed. The analysis results indicate that the stress condition, edge of etastic zone depth, supporting strength, and the friction angle and cohesion among coal stratum, roof and floor are sensitive factors. By introducing double-couple model, the layer dislocation rock burst was explained and the energy radiation characteristics were analyzed. The SOS micro-seismic monitoring system was applied to observe the rock burst hazards about a mining face. The results show that P- and S-wave energy radiations produced by rock burst have directional characteristics. The energy radiation characteristics of the 22 rock bursts occurring on 79Z6 long-wall face are basically the same as theoretical results, that is, the ratio of S-wave energy of sensor 4 to 6 is about 1.5 and that of P-wave is smaller than 0.5. The consistency of the monitored characteristics of the energy radiation theoretically increases with the total energy increasing.

Physical simulation of rock burst induced by stress waves

The behavior of stress wave propagation in rock walls and the process of rock bursts were simulated by application tests of material similar to rock. Results show that 1) the attenuation characteristics of stress waves were related to the material proper-ties, stress waves attenuate more quickly in soft material and 2) when the explosion load was applied at the top of the roadway, the number and the length of the cracks increased with a decrease in the distance between the explosive point and roof of the roadway. When the distance was 280 mm, only some chips appeared near the source, when the distance was 210 mm, some small cracks started to appear near the road-rib and when the distance was reduced to 140 mm, larger cracks appeared at the road-rib. It can be concluded that, under a given stress the number of cracks is closely related to the intensity of stress waves. The cracks in the sur-rounding rock can be reduced by controlling the intensity of the stress waves and rock bursts can be avoided to some extent by pre-venting the formation of layered crack structures. A new experimental approach has been provided for studying rock bursts by using physical simulation.

Frequency spectrum analysis on micro-seismic signal of rock bursts induced by dynamic disturbance

Blasting and breaking of hard roof are main inducing causes of rock bursts in coal mines with danger of rock burst,and it is important to find out the frequency spectrum distribution laws of these dynamic stress waves and rock burst waves for researching the mechanism of rock burst.In this paper,Fourier transform as a micro-seismic signal conversion method of amplitude-time character to amplitude-frequency character is used to analyze the frequency spectrum characters of micro-seismic signal of blasting,hard roof breaking and rock bursts induced by the dynamic disturbance in order to find out the difference and relativity of different signals.The results indicate that blasting and breaking of hard roof are high frequency signals,and the peak values of dominant frequency of the signals are single.However,the results indicate that the rock bursts induced by the dynamic disturbance are low frequency signals,and there are two obvious peak values in the amplitude-frequency curve witch shows that the signals of rock bursts are superposition of low frequency signals and high frequency signals.The research conclusions prove that dynamic disturbance is necessary condition for rock bursts,and the conclusions provide a new way to research the mechanism of rock bursts.

Gas-dynamic phenomena caused by rock mass tremors and rock bursts

Similar to coal, rock and gas ejections, rock mass tremors and rock bursts are among the most serious natural hazards accompanying the underground extraction of coal. Gas-dynamic phenomena caused by rock mass tremors and rock bursts observed as transient states of air parameters in mining headings,are usually generated as a result of a change in the geometry of headings and the release of considerable amounts of gases. Particular significance is attributed to transient states caused by disasters, which are often accompanied by rapid incidents, presenting threats to the life and health of the underground crew.In Polish mining there are known examples of transient states of air parameters recorded during gasdynamic phenomena, e.g. tremors and rock bursts. The paper presents the case studies of rapid seismic incidents to show how records in mine monitoring systems broaden the knowledge about the transient states of air parameters in mining headings generated because of them.

Controlling roof with potential rock burst risk through different pre-crack length:Mechanism and effect research

Roof pre-splitting is an effective method to control the roof with potential rock burst risk.In this study,three-point bending tests were carried out by using fine sandstone specimens with different pre-cracked lengths as test objects,and digital speckle correlation method(DSCM)and acoustic emission(AE)technology were used to track the entire process of crack propagation.The effect of pre-cracks on the fracture of rock beams was evaluated,and the mechanical mechanism of the rock beam fracture process was analyzed.The rock beam pre-splitting design method was developed,and the application effect of the method was proved by the microseismic monitoring data obtained from the 10303 working face of Jining No.2 coal mine in China.The results show that the loading time history curve of pre-cracked beams exhibits obvious residual characteristics.Compared with the intact rock beam,the tensile strength,and maximum tensile strain of 35 mm pre-cracked rock beam are decreased by 32.4% and 33.1%,respectively and the acoustic emission b value is increased by 30.2%.According to the pre-splitting design method of rock beam,the maximum and average microseismic energy of the 10303 working face after pre-splitting construction are reduced by 25.6% and 6.4%,respectively,with excellent prevention and control effect of thick roof.

Electromagnetic radiation early warning criterion of rock burst based on statistical theory

Rock bursts are serious natural disasters encountered worldwide in coal mining and rock engineering.In order to forecast rock bursts more effectively,a new rock burst forecasting index E,consisting of intensity and the number of pulses,is proposed,on the basis of abnormal characteristic symptoms of electromagnetic radiation(EMR) generated before rock bursts,combined with statistical theory.The index is distributed as a χ2 distribution with 2 degrees of freedom,i.e.,E～χ 2(2).Via this index,a quantitative comprehensive forecasting criterion of EMR was initially established.E values were calculated when the occurrence probability of the occurrence of a rock burst was 50%,70% and 90%.Appropriate measures should be taken when using these values on the scene.Using EMR data collected in the Nanshan Mine of the Hegang mining area,we verified that the analytical result were consistent with actual situations.This index is of theoretical importance and as a reference for forecasting rock bursts in coal mines.