Rock burst early warning method based on deformation localization and damage fracture mechanics

Abstract

The invention discloses a rock burst early warning method based on deformation localization and damage fracture mechanics, and the method comprises the steps: obtaining the micro-seismic data of each point of a to-be-early-warned region, the micro-seismic data comprising the three-dimensional coordinates and energy of the points; dividing the area to be pre-warned into a plurality of space areas by using a space grid algorithm, and gathering the micro-seismic data of each point of each space area into a point set; according to the micro-seismic data, the micro-seismic event gathering degree of each space area is judged and compared with a set degree threshold value, the space areas exceeding the degree threshold value are screened out, and the crack areas of the space areas are calculated; a crack tip factor is calculated according to the space region crack area, and the region stress of each space region can be obtained through the energy release rate and the damage deformation; according to the area stress of the space area and the measured critical stress, the impact danger coefficient is calculated, early warning is conducted on the space area according to the impact danger coefficient, and therefore the accuracy and efficiency of rock burst early warning are effectively improved.

Description

Technical Field

[0001] The invention relates to the technical field of mining engineering, and in particular to an impact rock pressure early warning method based on deformation localization and damage fracture mechanics. Background Art

[0002] In the current research, rock burst prediction methods mainly include microseismic monitoring, geoacoustic method, CT inversion method, electromagnetic radiation method and charge induction method. In actual engineering application scenarios, microseismic monitoring method has become the mainstream method widely adopted on site due to its high frequency of use and relatively accurate prediction effect. Due to the regional and nonlinear characteristics of rock burst in the process of incubation, traditional mechanical models are often difficult to accurately simulate this dynamic evolution process. In the mining process, most of them use energy field and stress field as indicators for dynamic monitoring. However, the monitoring range of stress field is small, and the energy field has the defect of insufficient mining of spatiotemporal characteristics of microseismic data. This leads to the problem of poor accuracy and low efficiency in the prediction of rock burst danger. Summary of the invention

[0003] The technical problem to be solved by the present invention is to provide a rock burst warning method based on deformation localization and damage fracture mechanics to address the deficiencies in the above-mentioned prior art, thereby effectively improving the accuracy and efficiency of rock burst warning.

[0004] In order to solve the above technical problems, the technical solution adopted by the present invention is: a rock burst early warning method based on deformation localization and damage fracture mechanics, comprising the following steps:

[0005] Step 1: Acquire microseismic data of each point in the area to be warned, wherein the microseismic data includes the three-dimensional coordinates of the point and the energy of the point;

[0006] Step 2: Divide the area to be warned into multiple spatial regions using a spatial grid algorithm, and collect the microseismic data of each point in each of the spatial regions into a point set;

[0007] Step 3: Determine the microseismic event aggregation degree of each of the spatial regions according to the microseismic data, compare it with a set degree threshold, and screen out the spatial regions that exceed the degree threshold;

[0008] Step 4: extract the high-density area of ​​the space area screened in step 3 to obtain the high-density area boundary, and then calculate the crack area of ​​the space area;

[0009] Step 5: Calculate the crack tip factor according to the crack area of ​​the spatial region, and screen out the spatial region that exceeds the factor threshold;

[0010] Step 6: Calculate the energy release rate and damage deformation energy of the spatial regions selected in step 5, and obtain the regional stress of each of the spatial regions through the energy release rate and damage deformation energy;

[0011] Step 7: Calculate the impact risk coefficient based on the regional stress of the space area and the measured critical stress, and issue an early warning to the space area based on the impact risk coefficient.

[0012] Furthermore, the warning area is a mining face or a mining tunnel.

[0013] Further, in step 2, the use of a spatial grid algorithm to divide the area to be warned into a plurality of spatial areas includes the following steps:

[0014] Determine the size of the window and the sliding step, and determine the number of the grids and the positions of the grids according to the size of the window;

[0015] Allocating the microseismic data to each of the grids according to the three-dimensional coordinates in the microseismic data;

[0016] The microseismic data in each of the grids is filtered.

[0017] Furthermore, filtering the microseismic data in each of the grids includes:

[0018] Determining the energy of each microseismic data in each of the grids, and removing the microseismic data with energy less than a set energy;

[0019] Counting the number of microseismic data in each of the grids;

[0020] Grids whose number of microseismic data is greater than a set number are screened out.

[0021] Further, in step 3, the degree of aggregation of microseismic events in each of the spatial regions is determined based on the microseismic data using Ripley's K function.

[0022] Further, in step 4, the high-density region is extracted from the spatial region screened out in step 3 to obtain the high-density region boundary through the KDE algorithm, and the crack area of ​​the spatial region is then calculated through the Monte Carlo algorithm.

[0023] Further, in step 7, the early warning of the space area according to the impact risk coefficient includes:

[0024] Determine the warning method according to the coefficient range of the impact risk coefficient;

[0025] An early warning is issued to the spatial area according to the early warning method.

[0026] Compared with the prior art, the present invention has the following advantages:

[0027] The present invention provides a rock burst early warning method based on deformation localization and damage fracture mechanics. Through the effective combination of algorithm and mechanics, it tracks the gestation and development process of microcracks inside coal and rock masses in real time, deeply analyzes the spatial evolution, temporal evolution, and energy evolution characteristics during the rupture process, achieves the goal of accurately predicting the occurrence of rock burst, better solves the complexity and nonlinear problems of rock burst, and can improve the accuracy and reliability of prediction in complex mining environments.

[0028] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 A schematic diagram of the overall process of a rock burst warning method based on deformation localization and damage fracture mechanics provided by the present invention.

[0030] Figure 2 Schematic diagram of spatial scanning and sliding window in the present invention.

[0031] Figure 3 Schematic diagram of kernel density estimation based on deformation localization in the present invention. DETAILED DESCRIPTION

[0032] like Figure 1-3 As shown, the present invention provides a rock burst warning method based on deformation localization and damage fracture mechanics, such as Figure 1 As shown, the following steps are included:

[0033] Step 1: Acquire microseismic data of each point in the area to be warned, wherein the microseismic data includes the three-dimensional coordinates of the point and the energy of the point;

[0034] Step 2: Divide the area to be warned into multiple spatial regions using a spatial grid algorithm, and collect the microseismic data of each point in each of the spatial regions into a point set;

[0035] Step 3: Determine the microseismic event aggregation degree of each of the spatial regions according to the microseismic data, compare it with a set degree threshold, and screen out the spatial regions that exceed the degree threshold;

[0036] Step 4: extract the high-density area of ​​the space area screened in step 3 to obtain the high-density area boundary, and then calculate the crack area of ​​the space area;

[0037] Step 5: Calculate the crack tip factor according to the crack area of ​​the spatial region, and screen out the spatial region that exceeds the factor threshold;

[0038] Step 6: Calculate the energy release rate and damage deformation energy of the spatial regions selected in step 5, and obtain the regional stress of each of the spatial regions through the energy release rate and damage deformation energy;

[0039] Step 7: Calculate the impact risk coefficient based on the regional stress of the space area and the measured critical stress, and issue an early warning to the space area based on the impact risk coefficient.

[0040] In the present invention, for rock burst warning, stress mechanics index is insensitive, and the existing method mainly predicts future impact events based on microseismic monitoring impact manifestation characteristics, which has a certain lag. In order to cope with the limitations of the current warning method, it is urgent to explore new theories and rock burst spatial area warning schemes based on intelligent algorithms. When the internal stress of the coal rock mass continues to concentrate and reaches the crack strength limit, the cracks expand rapidly and connect with each other, thus generating a local damage area. As the cracks continue to expand, the instability intensifies, the sound is released, and the accumulated elastic energy is violently released, causing rock burst to appear. The conditions for rock burst and deformation localization are consistent, so a new idea is proposed to move the prediction threshold forward, effectively combine the intelligent algorithm with the mechanical index, establish a strength criterion based on impact initiation, introduce the deformation localization method, and combine the relevant mechanical indicators of crack extension in fracture damage mechanics: energy release rate and damage strain energy, analyze the cumulative damage to predict rock burst, and use the determined impact hazard results to verify each other with the impact manifestation characteristics of actual microseismic monitoring, so as to optimize the judgment criteria. Finally, a rock burst warning method based on deformation localization and damage fracture mechanics was proposed, which effectively improved the accuracy and efficiency of rock burst warning.

[0041] In step 1, the warning area is a mining face or a mining tunnel. The microseismic data of the face or tunnel is read, and the data is filtered and converted to obtain a point set containing the three-dimensional coordinates and energy (x, y, z, m) of each point. Among them, (x, y, z) is the three-dimensional coordinate of the point, and m is the energy of the point.

[0042] In step 2, the use of a spatial grid algorithm to divide the area to be warned into multiple spatial areas includes the following steps:

[0043] Determine the size of the window and the sliding step, and determine the number of the grids and the positions of the grids according to the size of the window;

[0044] Allocating the microseismic data to each of the grids according to the three-dimensional coordinates in the microseismic data;

[0045] The microseismic data in each of the grids is filtered.

[0046] The filtering of the microseismic data in each of the grids includes:

[0047] Determining the energy of each microseismic data in each of the grids, and removing the microseismic data with energy less than a set energy;

[0048] Counting the number of microseismic data in each of the grids;

[0049] Grids whose number of microseismic data is greater than a set number are screened out.

[0050] In step 2, the spatial scanning algorithm is used to divide the grid window, and the microseismic events are assigned to the corresponding grids according to their coordinates and time. Then, the microseismic spatial data are projected onto the plane to complete the division of multiple different spatial areas of the working face.

[0051] The spatial scanning algorithm is used to divide the grid window, determine the spatial window size and the number of sliding grids, and distribute the microseismic events to the corresponding grids according to their coordinates and time, thereby forming microseismic data sets in different spatial regions. Figure 2 shown.

[0052] The spatial scanning method is as follows:

[0053] 1) First, determine the scope of the spatial scan and set the length and width of the working surface rectangle to accommodate all microseismic event points to form an enclosing grid.

[0054] 2) Determine the size of the grid window and the sliding step size. Determine the number of grids and the coordinate position of each grid according to the length and width of the enclosing grid.

[0055] 3) Allocate event points to corresponding grids. Filter data points one by one, determine whether they are within the enclosing grid range based on their coordinate positions, and then transfer the point information (including date, coordinates, and energy) to the corresponding grid. Finally, aggregate to form grid data point sets in different areas of the working face or laneway.

[0056] In step 3, the degree of microseismic event aggregation in each of the spatial regions is determined based on the microseismic data using Ripley's K function.

[0057] Step 3: Calculate Ripley's K function for the data point set in each spatial region, set the K function threshold, and when the degree of spatial microseismic event aggregation exceeds the specified threshold, it means that the region is strongly affected by dynamic load disturbance and deformation localization is initiated. If it does not exceed the threshold, calculate the K function of the next spatial region.

[0058] The deformation localization method in the present invention refers to the phenomenon that significant plastic deformation occurs in certain areas of the material during the stress process, resulting in increased local strain. It is usually used to describe the behavior of materials under complex loading conditions, especially in the yield and fracture analysis of materials. Its basic principles are based on the following points:

[0059] 1) Strain concentration: Under a specific stress state, local stress concentration areas will form inside the material, and the deformation of these areas is significantly greater than the overall deformation.

[0060] 2) Destruction mechanism: When the stress of coal rock mass approaches the strength limit, a strain concentration area will be generated, and the microcracks will further expand to form macro cracks. When the macro cracks expand to an unstable state, the elastic energy is released to form an impact, and then the system rebalances, and the crack expansion rate slows down or stops.

[0061] Ripley's K function is a method for detecting and quantifying the spatial clustering of point patterns, which is suitable for spatial analysis at different scales. It measures the degree of clustering by counting the number of points within an arbitrary distance r. The threshold can be determined by a significance test. Multiple samples are generated by simulating the K function under a completely random distribution, and its 95% confidence interval is calculated. Then, it is determined whether the actual K function exceeds this range. If it exceeds, it can be regarded as a significant clustering, and the part exceeding the range can be used as a threshold. At the same time, in multi-scale analysis, the threshold can be dynamically adjusted as the scale changes. For example, local clustering is found at a small scale, and then whether there is a more extensive clustering phenomenon in a larger area at a large scale. Adjusting the threshold can help more accurately identify the localized characteristics of deformation at different scales. Thereby quantitatively evaluating the clustering characteristics of microseismic events in different spatial regions. Its basic principles are mainly as follows:

[0062] Principle: The K function calculates the number of point pairs within a given radius r and compares it with the expected number of point pairs.

[0063] formula:

[0064]

[0065] Where A is the area of ​​the study area, n is the number of points, and d ij is the distance between points i and j, I(d ij ≤r) is an indicator function, which takes the value of 1 when the distance between i and j is less than or equal to r, and 0 otherwise.

[0066] In step 4, the high-density region is extracted from the spatial region selected in step 3 to obtain the high-density region boundary through the KDE algorithm, and the crack area of ​​the spatial region is then calculated through the Monte Carlo algorithm.

[0067] The present invention uses the KDE algorithm to set the density threshold for the deformation localization start-up area, generates a kernel density function, and extracts the high-density area boundary. The high-density area boundary is obtained according to the KDE algorithm, and the Monte Carlo algorithm is used to obtain the estimation of the crack area in the area.

[0068] The KDE (Kernel Density Estimation) kernel algorithm is a non-parametric method for estimating the probability density function of a random variable. The specific steps of the present invention are: regional microseismic data - kernel density estimation (selecting kernel function and bandwidth parameters, estimating the kernel density of each point and marking the generated density map) - determining high-density areas (setting density thresholds, extracting high-density areas). The kernel density estimation based on deformation localization is shown in the figure below: Figure 3 As shown, the basic principles are as follows:

[0069] Kernel function: KDE uses a kernel function (such as a Gaussian kernel) to smooth the data points. The influence of each data point is distributed to the surrounding area through the kernel function.

[0070] Density estimation: By summing the kernel functions of all data points, a smooth estimate of the data distribution can be obtained. The formula is The estimated probability density function represents the density estimate at point x. n is the number of samples; K is the Gaussian kernel function; x i is the value of the i-th sample point, x is the point whose density is expected to be estimated; h is the bandwidth parameter, which controls the smoothness of the kernel function.

[0071] The Monte Carlo algorithm is a method for numerical calculation using random sampling. The specific steps of the present invention are: define the region boundary (i.e., extract the high-density region using the KDE algorithm) - generate random points (generate N sample points to cover the defined region) - estimate the crack area (calculate the high-density region). The basic principle is as follows:

[0072] 1) Random sampling: Simulating complex systems or processes by randomly generating a large number of samples.

[0073] 2) Estimation: Estimate the quantity of interest (such as integral, expected value, etc.) based on the characteristics and statistical laws of the sample. Common applications include financial models, physical simulations, etc.

[0074] Convergence: As the number of samples increases, Monte Carlo estimates usually converge to the true value according to the law of large numbers.

[0075] In step 5-7, the crack tip factor K of the area is calculated based on the relevant theories and mechanical formulas of damage fracture mechanics. I If the value exceeds the K measured in the laboratory IThreshold, indicating that the crack in the region further expands and evolves, which may induce rock burst, so the energy release rate and damage deformation energy are calculated, and finally the regional stress σ in the region is obtained. 区域 Otherwise, the next spatial area is calculated. Sampling and testing of coal and rock mass in the mining impact area is carried out, and the critical stress σ of the current assessment area is calculated based on damage fracture mechanics and laboratory measured data. 临界 , if the regional stress σ 区域 Greater than the critical stress σ 临界 , it means that the impact risk is high, and if it is less than, it means that the impact risk is not high.

[0076] The critical stress σ in the present invention 临界 Measured in the laboratory.

[0077] In the present invention, understanding the mechanical mechanism of rock burst is the key to successful early warning. However, the sensitivity of mechanical indicators to rock burst prediction and early warning is low. The current rock burst early warning method is mainly based on microseismic or ground sound perception, and the monitoring and analysis are relatively lagging. It fails to effectively introduce mechanical indicators into the monitoring and early warning of rock burst, and cannot effectively evaluate the impact hazard.

[0078] To address this problem, the relevant mechanical indicators of crack extension in the fracture damage mechanics theory are introduced to effectively combine the intelligent algorithm with the mechanical indicators. Due to the complexity of rock burst, considering the influence of the extension of macro cracks and cracks, damage to coal and rock materials occurs, the continuous extension of cracks leads to damage accumulation, localized deformation starts, and rock burst begins to breed. Fracture mechanics and damage mechanics can be organically combined, and intelligent algorithms can be used to evaluate the degree of crack extension and analyze the accumulated damage to predict rock burst. The determined rock burst hazard results are verified with the actual impact manifestation characteristics of microseismic monitoring to optimize the determination criteria. The following is the solution to the impact hazard coefficient ρ:

[0079] Based on the deformation localization method, the crack area A in a certain spatial region is estimated. Id , has the following relationship with the released energy:

[0080]

[0081] Where dE is the energy released during crack growth, and dA is the surface area increased by crack growth, in m 2 If the surface area increment dA of crack extension is known, the formula can be approximated as:

[0082]

[0083] Therefore, the energy release rate G of the crack can be obtained by statistically obtaining the total energy E of a certain area and the estimated crack area A. IdAccording to the elastic modulus of the damage variable, the damage deformation energy and the energy release rate, the ultimate strength of the damage fracture G can be obtained. Ic , and further obtain the current regional stress σ 区域 At the same time, sampling and testing of coal and rock mass in the mining impact area are carried out to calculate the critical stress σ of the current assessment area. 临界 Therefore, the impact risk coefficient ρ can be calculated from the regional stress σ 区域 Specific critical stress σ 临界 Find:

[0084]

[0085] The degree of impact risk is comprehensively evaluated by the obtained impact risk coefficient, that is, the impact risk coefficient. When the regional stress is greater than the critical stress, there is an impact risk. When the regional stress is less than or equal to the critical stress, there is no impact risk.

[0086] In general, in the present invention, the regional stress is obtained by the above-mentioned calculation method, and the critical stress is obtained by specific laboratory measurements.

[0087] Therefore, the early warning of the space area according to the impact risk coefficient includes:

[0088] Determine the warning method according to the coefficient range of the impact risk coefficient;

[0089] An early warning is issued to the spatial area according to the early warning method.

[0090] That is, when the impact risk coefficient is less than or equal to 1, there is no impact risk and no warning is issued. When the impact risk coefficient is greater than 1, there is an impact risk and a warning is issued for the spatial area.

[0091] In the present invention, rock burst has certain regional characteristics and nonlinear characteristics, and is easily affected by geological conditions and noise. The current monitoring and early warning methods have poor accuracy and reliability in spatial prediction, and fail to find more accurate spatial danger areas. It is difficult to reflect the precursor characteristics and significant spatial characteristics of the impact, and the judgment of the regional danger level is also vague.

[0092] In response to this problem, the deformation localization theory is introduced. When the coal-rock fracture area is subjected to a large disturbance, deformation is likely to occur, which is manifested as an increase in microseismic events in microseismic monitoring. According to historical experience, rock bursts often appear in a certain area. Localized deformation and damage is the basic law that causes deformation and damage of surrounding rock structures. The internal stress of the coal body reaches the limit, and the cracks continue to expand and penetrate until the stored elastic energy is violently released in an instant, which is consistent with the conditions for rock bursts. The process of localized deformation of rock bursts can be characterized based on the on-site monitoring data of the working face or tunnel.

[0093] Therefore, for the impact hazard assessment of a region, it is first necessary to determine whether it has the characteristics of local destruction. The present invention first uses Ripley's K function to determine the spatial aggregation of events based on the microseismic events in the divided area. This function can calculate the aggregation under different scales r. First, a significance threshold is set. When the K function value in a certain spatial area does not exceed the threshold, the localization of coal and rock deformation has not started, and the incubation and evolution process of rock burst has not yet begun. When the K value exceeds the threshold, it means that there is an aggregation phenomenon within a certain specific radius. At this time, it is considered that the spatial distribution of microseismic events is concentrated and the localization of deformation is started.

[0094] After the localization of regional deformation is started, the density function is generated by the KDE algorithm, the kernel density value of the region is estimated, and the boundary of the high-density area is extracted. It is difficult to calculate the area using the integral method. The Monte Carlo method is used to generate random points to cover the high-density area defined by the KDE algorithm, and the crack area A is estimated. Then, the degree of impact danger is quantitatively evaluated in combination with the mechanical indicators of fracture damage theory. This method can effectively obtain the mechanical precursor characteristics of rock burst, advance the evaluation of the degree of danger of rock burst incubation and evolution, effectively move the prediction threshold forward, and improve the accuracy and reliability of spatial prediction of rock burst.

[0095] The present invention combines intelligent algorithms to perform real-time intelligent analysis of monitoring data, quickly realizes quantitative calculation of deformation degree in different areas through spatial scanning ideas, evaluates the expansion of surrounding rock cracks and the degree of surrounding rock damage in different areas, fully mines the geophysical information of microseismic monitoring data, tracks the localized evolution characteristics of deformation, and timely analyzes the precursor characteristics and evolution laws of rock burst. Based on the existing mechanical parameter data and the monitoring records of the on-site microseismic database, the impact hazard degree is calculated in real time, batch processing of monitoring data is automatically realized, and early warning information is analyzed and given in a timely manner, thereby minimizing the delay of microseismic monitoring and analysis and improving the timeliness and reliability of predicting rock burst risks.

[0096] The above description is only a preferred embodiment of the present invention and does not limit the present invention in any way. Any simple modification, change and equivalent structural change made to the above embodiment based on the technical essence of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. A rock burst early warning method based on deformation localization and damage fracture mechanics, characterized in that: The following steps are involved: Step 1: Acquire microseismic data of each point in the area to be warned, wherein the microseismic data includes the three-dimensional coordinates of the point and the energy of the point; Step 2: Divide the area to be warned into multiple spatial regions using a spatial grid algorithm, and collect the microseismic data of each point in each of the spatial regions into a point set; Step 3: Determine the microseismic event aggregation degree of each of the spatial regions according to the microseismic data, compare it with a set degree threshold, and screen out the spatial regions that exceed the degree threshold; Step 4: extract the high-density area of ​​the space area screened in step 3 to obtain the high-density area boundary, and then calculate the crack area of ​​the space area; Step 5: Calculate the crack tip factor according to the crack area of ​​the spatial region, and screen out the spatial region that exceeds the factor threshold; Step 6: Calculate the energy release rate and damage deformation energy of the spatial regions selected in step 5, and obtain the regional stress of each of the spatial regions through the energy release rate and damage deformation energy; Step 7: Calculate the impact risk coefficient based on the regional stress of the space area and the measured critical stress, and issue an early warning to the space area based on the impact risk coefficient.

2. A rock burst early warning method based on deformation localization and damage fracture mechanics according to claim 1, characterized in that: The warning area is a mining face or a mining tunnel.

3. A rock burst early warning method based on deformation localization and damage fracture mechanics according to claim 1, characterized in that: In step 2, the use of a spatial grid algorithm to divide the area to be warned into multiple spatial areas includes the following steps: Determine the size of the window and the sliding step, and determine the number of the grids and the positions of the grids according to the size of the window; Allocating the microseismic data to each of the grids according to the three-dimensional coordinates in the microseismic data; The microseismic data in each of the grids is filtered.

4. A rock burst warning method based on deformation localization and damage fracture mechanics according to claim 2, characterized in that: The filtering of the microseismic data in each of the grids includes: Determining the energy of each microseismic data in each of the grids, and removing the microseismic data with energy less than a set energy; Counting the number of microseismic data in each of the grids; Grids whose number of microseismic data is greater than a set number are screened out.

5. The rock burst warning method based on deformation localization and damage fracture mechanics according to claim 1, characterized in that: In step 3, the degree of microseismic event aggregation in each of the spatial regions is determined based on the microseismic data using Ripley's K function.

6. A rock burst early warning method based on deformation localization and damage fracture mechanics according to claim 1, characterized in that: In step 4, the high-density region is extracted from the spatial region selected in step 3 to obtain the high-density region boundary through the KDE algorithm, and the crack area of ​​the spatial region is then calculated through the Monte Carlo algorithm.

7. The rock burst warning method based on deformation localization and damage fracture mechanics according to claim 1, characterized in that: In step 7, the step of issuing an early warning to the space area according to the impact risk coefficient includes: Determine the warning method according to the coefficient range of the impact risk coefficient; An early warning is issued to the spatial area according to the early warning method.

Images