Rock burst early warning method based on multivariate catastrophe information fusion

Abstract

The invention discloses a rockburst early warning method based on multivariate catastrophe information fusion, and belongs to the field of underground engineering rockburst prediction.The method comprises the steps that a standard rock sample is placed in a true triaxial test system, the rockburst inoculation process is simulated through radial rapid unloading and tangential continuous loading, and test basic data are synchronously collected; performing normalization processing on the test basic data to obtain a normalized data set with unified dimensions; establishing a fusion function model based on the normalized data set, and determining a key stage division result of the rockburst inoculation process by analyzing time sequence distribution characteristics of extreme points and inflection points of the fusion function model; and based on the key stage division result of the rockburst inoculation process, constructing a rockburst early warning chain from the internal state, the external performance, the time sequence and the spatial distribution of rocks, and realizing early warning of rockburst disasters.

Description

Technical Field

[0001] The present invention belongs to the technical field of underground engineering rockburst prediction, and in particular relates to a rockburst early warning method based on multi-dimensional disaster information fusion. Background Art

[0002] To prevent rockbursts and the resulting hazards, numerous researchers have proposed a variety of rockburst prediction methods based on stiffness theory, strength theory, energy theory, fractal theory, catastrophe theory, deformation instability theory, and impact tendency theory, significantly advancing the development of rockburst propensity prediction. However, the results of uniaxial, biaxial, and conventional triaxial rockburst tests cannot accurately reflect the three-dimensional stress state changes during rockburst occurrence, making it difficult to effectively simulate the entire rockburst incubation, occurrence, development, and destruction process. Therefore, conducting true triaxial tests involving rapid radial unloading, three-dimensional five-sided loading, and continuous tangential loading can address the shortcomings of previous laboratory rockburst tests and provide support for the development of more accurate rockburst early warning methods based on multi-dimensional disaster information fusion, thereby mitigating or delaying rockburst occurrence. Furthermore, the hazard source (energy release body - rockburst body) undergoes dynamic changes throughout the rockburst incubation and evolution process, and its multi-physics field monitoring information is the foundation of rockburst early warning. However, due to the complex construction conditions of underground projects, ideal monitoring environments are not available. Consequently, multi-dimensional information on the entire evolution of rockburst hazards is difficult to obtain on-site. Furthermore, rockburst early warning methods applied on-site often struggle to fully implement fusion analysis and comprehensive early warning based on multiple monitoring information. Furthermore, existing experimental results have mostly focused on multi-dimensional rockburst monitoring, without delving deeply into the correlation and fusion analysis of multi-dimensional monitoring information. Consequently, a multi-information fusion early warning method suitable for rockburst disasters has yet to be established. Therefore, this paper proposes a rockburst early warning method based on the fusion of multi-dimensional disaster information. Summary of the Invention

[0003] In order to solve the above technical problems, the present invention proposes a rockburst early warning method based on multi-dimensional disaster information fusion to solve the problems existing in the above-mentioned prior art.

[0004] To achieve the above objectives, the present invention provides a rockburst early warning method based on multi-dimensional disaster information fusion, comprising:

[0005] The standard rock specimens were placed in a true triaxial test system, and the rockburst incubation process was simulated through radial rapid unloading and tangential continuous loading, while basic test data was collected simultaneously.

[0006] Normalizing the experimental basic data to obtain a normalized data set with unified dimensions;

[0007] Establishing a fusion function model based on the normalized data set, and determining the key stage division results of the rockburst incubation process by analyzing the time series distribution characteristics of the extreme points and inflection points of the fusion function model;

[0008] Based on the key stage division results of the rockburst incubation process, a rockburst early warning chain is constructed from the internal state, external performance, time series and spatial distribution of the rock to achieve early warning of rockburst disasters.

[0009] Optionally, the process of collecting basic test data includes:

[0010] Based on the same piece of granite, rectangular prisms were prepared to obtain standard rock samples;

[0011] A fully digital servo controller is used to control the independent loading and unloading in the vertical and horizontal directions of the true triaxial test system, and the test is carried out according to the radial rapid unloading and tangential continuous loading path; at the same time, the acquisition equipment synchronously collects multi-dimensional disaster information and transmits the multi-dimensional disaster information in real time to the computer for storage and preliminary processing to obtain basic test data.

[0012] Optionally, the experimental basic data are processed based on a deviation standardization method to obtain a normalized data set with unified dimensions.

[0013] Optionally, the process of establishing a fusion function model includes:

[0014] The normalized data were modeled based on the rational number function fitting method to obtain the fitting functions of stress, acoustic emission, total energy and dissipated energy;

[0015] Performing parameter fitting on the fitting function to obtain a fitting curve;

[0016] An optimized fusion function model is obtained based on the fitting curve.

[0017] Optionally, the fusion function model is:

[0018]

[0019] Where, f i (x) is the fusion function model, m and n are the orders of the highest order x term in the denominator and numerator respectively; f1(x) is the stress function, f2(x) is the acoustic emission function, f3(x) is the total energy function, and f4(x) is the dissipated energy function; p ij ,q ij is the coefficient of the x term.

[0020] Optionally, the process of determining the key stage division result of the rockburst incubation process includes:

[0021] Based on the first-order derivative and second-order derivative of the fusion function model, the intersection of the derivative function curve and the time axis is analyzed to determine the location of the extreme points and inflection points;

[0022] Determine the dividing point between the pre-disaster period and the post-disaster period based on the positions of the extreme points and the inflection points;

[0023] Based on the changing law of the function's concavity and convexity, the critical points between the rockburst development period and the impending disaster period are obtained.

[0024] The key stage division results of the rockburst incubation process are determined based on the dividing point between the pre-disaster period and the post-disaster period and the critical point between the rockburst development period and the pre-disaster period.

[0025] Optionally, the process of building a rockburst early warning chain includes:

[0026] Based on the key stage division results of the rockburst incubation process, stress precursor characteristics are obtained through extreme point analysis of the stress function;

[0027] Based on the extreme point analysis of the acoustic emission function, the acoustic emission precursor characteristics are obtained;

[0028] Based on the inflection point analysis of the total energy function and the dissipated energy function, the energy precursor characteristics are obtained;

[0029] Establishing an early warning mechanism based on the stress precursor characteristics, acoustic emission precursor characteristics and energy precursor characteristics;

[0030] Based on the real-time monitoring data of the specimen failure morphology, the failure precursor characteristics of the unloading surface splitting into slabs and the bending of the rock slab are obtained, and a late warning mechanism is established based on the failure precursor characteristics;

[0031] Through the temporal and spatial association of the early warning mechanism and the late warning mechanism, a rock burst warning chain is constructed.

[0032] Compared with the prior art, the present invention has the following advantages and technical effects:

[0033] The rockburst early warning method based on the fusion of multi-dimensional disaster information of the present invention simulates the rockburst incubation process through a true triaxial test system, synchronously collects multi-dimensional disaster information such as stress, strain, and acoustic emission, and normalizes the data, eliminating the dimensional differences of different physical quantities, making the data more comparable and valuable for analysis. By establishing a fusion function model and analyzing its extreme points and inflection points, the key stages of rockburst incubation are accurately divided, providing a scientific basis for early warning. Further, a rockburst early warning chain is constructed from four aspects: the internal state of the rock, external manifestations, time series, and spatial distribution, realizing real-time and continuous early warning of rockburst disasters from multiple dimensions. This method can directly, in real time, and continuously obtain comprehensive information on the entire process of rockburst incubation and development, effectively reducing the risk of engineering accidents that may be caused by rockbursts, and providing strong technical support for the early warning and prevention of rockburst disasters in underground engineering. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] The accompanying drawings, which constitute part of this application, are intended to provide a further understanding of this application. The exemplary embodiments and descriptions of this application are intended to explain this application and do not constitute an improper limitation on this application. In the accompanying drawings:

[0035] Figure 1 1 is the stress-strain curve of the rock sample under different tangential loading rates according to the embodiment of the present invention;

[0036] Figure 2 The time series changes of the acoustic emission impact number and the cumulative acoustic emission impact number at 1.0 MPa / s in the embodiment of the present invention;

[0037] Figure 3 The time series changes of the acoustic emission impact number and the cumulative acoustic emission impact number at 3.0 MPa / s in the embodiment of the present invention;

[0038] Figure 4 The time series changes of the acoustic emission impact number and the cumulative acoustic emission impact number at 5.0 MPa / s in the embodiment of the present invention;

[0039] Figure 5 The changing characteristics of total energy, elastic strain energy and dissipated energy at peak stress of rock samples under different tangential loading rates in the embodiment of the present invention;

[0040] Figure 6 The particle size grouping and mass distribution characteristics of the fragments at 1.0 MPa / s in the embodiment of the present invention;

[0041] Figure 7 The particle size grouping and mass distribution characteristics of the fragments at 3.0 MPa / s in the embodiment of the present invention;

[0042] Figure 8 The particle size grouping and mass distribution characteristics of the fragments at 5.0 MPa / s in the embodiment of the present invention;

[0043] Figure 9 This is the fusion analysis of multi-dimensional rockburst disaster information in an embodiment of the present invention;

[0044] Figure 10 This is a rockburst multivariate information fitting function curve according to an embodiment of the present invention;

[0045] Figure 11 : This is a rockburst multivariate information derivative function curve according to an embodiment of the present invention, where (a) is the first-order derivative and (b) is the second-order derivative;

[0046] Figure 12 The evolution state of rock burst under 1.0 MPa / s, where (a) is the stress precursor, (b) is the acoustic emission precursor, (c) is the energy precursor, (d) is the destruction precursor, and (e) is the evolution stage based on normalized fusion analysis;

[0047] Figure 13The evolution state of rock burst under 3.0 MPa / s, where (a) is the stress precursor, (b) is the acoustic emission precursor, (c) is the energy precursor, (d) is the destruction precursor, and (e) is the evolution stage based on normalized fusion analysis;

[0048] Figure 14 The evolution state of rock burst under 5.0 MPa / s, where (a) is the stress precursor, (b) is the acoustic emission precursor, (c) is the energy precursor, (d) is the destruction precursor, and (e) is the evolution stage based on normalized fusion analysis;

[0049] Figure 15 This is a schematic diagram of a rockburst precursor information chain according to an embodiment of the present invention;

[0050] Figure 16 This is an overall flow chart of an embodiment of the present invention. DETAILED DESCRIPTION

[0051] It should be noted that, in the absence of conflict, the embodiments and features of the embodiments in this application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0052] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0053] Example 1

[0054] like Figure 16 As shown, this embodiment provides a rockburst early warning method based on multivariate disaster information fusion, comprising the following steps: placing a standard rock specimen in a true triaxial test system, simulating the rockburst incubation process by radial rapid unloading and tangential continuous loading, and synchronously collecting basic test data; normalizing the basic test data of the rockburst early warning method based on multivariate disaster information fusion to obtain a normalized data set of unified dimension; establishing a fusion function model based on the normalized data set, and determining the key stage division result of the rockburst incubation process by analyzing the time series distribution characteristics of the extreme points and inflection points based on the fusion function model; based on the key stage division result of the rockburst incubation process, constructing a rockburst early warning chain from the internal state, external performance, time series and spatial distribution of the rock to realize the early warning of rockburst disasters.

[0055] Step 1: All samples were taken from the same complete granite rock and processed into rectangular prism samples with a size of 100 mm (length) × 100 mm (width) × 200 mm (height).

[0056] Step 1-1: All specimens were taken from a single, intact granite rock. After processing, the specimens were 100 mm long and wide, and 200 mm high. Sample preparation strictly adhered to the International Association of Rock Mechanics standards, with a flatness tolerance of ±0.05 mm between opposing surfaces and a perpendicularity tolerance of ±0.25° between adjacent surfaces. All sides and ends of the specimens were finely ground and polished to minimize local stress concentrations and produce smooth, flat end faces. Prior to loading, both ends of the specimen loading surface were coated with a grease lubricant and padded with filter paper, followed by preloading to minimize noise signals generated by friction at the ends.

[0057] Step 2: Conduct true triaxial tests on rock specimens and equip them with a real-time collection and monitoring system for multi-dimensional disaster information. Further process, analyze, and calculate the basic test data obtained to obtain multi-dimensional disaster information such as the ejection failure characteristics, stress, strain, energy, and acoustic emission characteristics of the rock specimens.

[0058] Step 2-1: Using a true triaxial rockburst test system, conduct true triaxial rockburst tests under conditions of radial rapid unloading and tangential loading at varying rates. This realistically recreates the entire rockburst initiation and evolution process and analyzes the multi-dimensional catastrophic information evolution characteristics of the rockburst initiation process. The true triaxial rockburst test system, controlled by a fully digital servo-controlled measurement and control system, enables independent loading and unloading in both vertical and horizontal directions, with maximum loading pressures of 5000 kN and 3000 kN in the vertical and horizontal directions, respectively. This system provides real-time acquisition of force and displacement throughout the rockburst test, ensuring smooth control of the test progress. Furthermore, the testing machine is equipped with a high-precision servo-controlled single-sided rocker unloading device, enabling independent loading and unloading in three vertical directions without interfering with each other. During the true triaxial rockburst test, observation equipment such as acoustic emission and high-speed digital cameras are used to synchronously acquire and process multiple data signals in real time. After monitoring by sensors, the data is transmitted to a multi-channel acquisition card for display and storage by computer software. This enables the simultaneous acquisition and processing of multiple data channels in real time, yielding a complete temporal evolution curve.

[0059] Furthermore, the process of collecting basic test data includes: obtaining a standard rock sample based on the same piece of rectangular prism granite; using a fully digital servo measurement and controller to control the independent loading and unloading in the vertical and horizontal directions of the true triaxial test system, and conducting tests according to radial rapid unloading and tangential continuous loading paths; at the same time, the acquisition equipment synchronously collects multi-dimensional disaster information, and transmits the multi-dimensional disaster information in real time to the computer for storage and preliminary processing to obtain basic test data.

[0060] Step 2-2: The stress-strain curves of the rock samples under different tangential loading rates are as follows: Figure 1As shown in the figure. (a) The failure process of the rock sample under different tangential loading rates roughly goes through four stages: initial compaction stage (OA stage), elastic deformation stage (AB stage), pre-peak unstable fracture stage (BC stage), and post-peak failure ejection stage (CD stage). (b) In the OA stage, a slight concave trend is observed, with microcracks within the sample beginning to close and initiate, and localized particle ejection occurs on the unloading surface, but no obvious fracture occurs. (c) In the AB stage, a good linear relationship is observed, with the slope of the curve being roughly the same, indicating that the elastic moduli of each group of samples are similar and their average values ​​are good. Rock plate fracture and particle ejection begin to occur on the unloading surface. (d) In the BC stage, a clear yield plateau appears. The values ​​of the yield point and peak point increase with increasing tangential loading rate, and the curve between the two points changes gently. Plate fracture and fragment ejection occur on the unloading surface. (e) The corresponding failure characteristics in the CD stage are comprehensive fragment ejection on the unloading surface and tensile-shear failure within the potential rockburst crater. The stress-strain curve drops sharply and is relatively steep.

[0061] The temporal changes of the acoustic emission impact number and the cumulative acoustic emission impact number under different tangential loading rates are shown in Figure 2. Figure 2 、 Figure 3 、 Figure 4 As shown. Figure 2 、 Figure 3 、 Figure 4 The acoustic emission characteristics of each stage are summarized as follows: (a) Stage I: There is a small amount of acoustic emission activity. The number of AE impacts for each rock sample begins to increase, then fluctuates within a certain range. The number of microcracks increases slightly. The rock sample subjected to a tangential loading rate of 5.0 MPa / s exhibits particularly active acoustic emission activity. During this stage, the AE signals of each rock sample remain at a low level, and acoustic emission activity is inactive. The influence of different tangential loading rates on the acoustic emission evolution characteristics of each rock sample is not significant. (b) Stage II: The number of AE impacts increases, and cracks within the rock sample rapidly expand and continue to grow at a high rate. In the latter part of this stage, the increase is greatest for the rock sample subjected to a tangential loading rate of 5.0 MPa / s, and the increases decrease in the rock samples subjected to tangential loading rates of 3.0 or 1.0 MPa / s. (c) Stage III: The number of AE impacts for each rock sample first increases to a maximum level, then rapidly decreases. During this stage, potential rockburst craters gradually develop within the rock sample, and damage phenomena such as particle ejection and plate breakage and spalling occur on the unloaded surface. When the vertical stress exceeds the peak stress of the rock sample, the degree of microcrack clustering increases dramatically, leading to the formation of macroscopic fracture surfaces in the rock. (d) Stage IV: The rock sample collapses and loses its bearing capacity. All rock samples experience a quiet period with weak acoustic emission activity, and the AE impact number reaches a minimum.

[0062] The changing characteristics of total energy, elastic strain energy and dissipated energy at peak stress of rock samples under different tangential loading rates are as follows: Figure 5 As shown. Figure 5 It can be seen that (a) the elastic strain energy curve shows a sharp increase followed by a slow rise with increasing tangential loading rate; the overall trend of the total energy and dissipated energy curves is relatively similar, showing a sharp increase followed by a brief decrease and finally a sharp increase, with the demarcation point roughly within the range of 1 to 3 MPa / s. (b) Analysis of the characteristic energy peaks shows that the elastic strain energy accumulation and dissipated energy release capabilities are closely related to the tangential loading rate effect. As the tangential loading rate increases, the specimen accumulates energy while dissipating energy. More external energy input and elastic strain energy release are converted into kinetic energy of the broken rock, resulting in rockburst and severe damage. In addition, the higher the tangential loading rate, the more obvious the fracture traces remaining on the specimen failure surface and the more fully developed the through cracks, the more pronounced the brittle failure characteristics, and the greater the failure strength. (c) At different tangential loading rates, dissipated energy is significantly smaller than elastic strain energy. The lower the tangential loading rate, the slower the growth rate of microcracks within the specimen, and the longer the damage accumulation time, resulting in a higher degree of degradation within the specimen. Increasing the tangential loading rate accelerates the growth rate of microcracks and reduces their growth time, preventing them from developing and expanding fully within a short period of time. Therefore, increasing the tangential loading rate reduces the degree of damage degradation within the rock, which is the fundamental reason why the peak elastic strain energy and dissipated strain energy vary regularly with the tangential loading rate.

[0063] The particle size grouping and mass distribution characteristics of fragments under different tangential loading rates are as follows: Figure 6 、 Figure 7 、 Figure 8 As shown. Figure 6 、 Figure 7 、 Figure 8 As can be seen, (a) the mass of rockburst fragments is primarily composed of coarse (d ≥ 9.5 mm), medium (4.75 mm ≤ d < 9.5 mm), and fine (0.075 mm ≤ d < 4.75 mm) fragments. Coarse fragments have the largest mass percentage, followed by fine and medium fragments, and fine (d < 0.075 mm) fragments have the smallest mass percentage. Furthermore, coarse fragments mostly exhibit irregular flakes and thin wedge-shaped structures, while medium and fine fragments are primarily ridged and blocky, and fine fragments are primarily powdery. This indicates that changes in the tangential loading rate cause changes in the mass percentage of fragments of different particle size groups. With increasing tangential loading rate, the fraction of fine fragments remains essentially unchanged, while the fractions of fine, medium, and coarse fragments initially increase and then decrease. (b) With increasing tangential loading rate, the mass of fragments of various particle size groups and the total mass of fragments generated during the entire rock failure process gradually increase. The results show that the greater the tangential loading rate, the greater the degree of surrounding rock damage caused by the rockburst; and a large amount of energy will be released instantly when the rockburst occurs, causing serious damage to the rock sample and dissipating a lot of energy, resulting in a large number of coarse-grained, medium-grained and fine-grained fragments.

[0064] Step 3: Analyze the evolution characteristics of the multivariate disaster information during the rockburst incubation process, normalize the changing trend of the multivariate disaster information, and clarify the correlation effect between the multivariate disaster information during the rockburst incubation process.

[0065] Step 3-1: The rockburst physical simulation test collected a large amount of multi-physical field information such as stress, energy and acoustic emission, but the value ranges of different physical quantities vary greatly and the units are inconsistent. Therefore, the data of different parameters cannot be plotted in the same coordinate system, which is not conducive to the macro-analysis of the changing trends of various physical quantities and their mutual influence. In order to eliminate the differences in the values ​​and dimensions of various physical quantities, it is necessary to normalize the multivariate information collected in the test. The data after normalization are all within the range of 0 to 1, which eliminates the influence of dimension while highlighting the changing laws of various physical quantities. Taking into account the discreteness of the test collected data and the convenience of data processing, this embodiment uses the deviation normalization method to normalize the test data, and assumes that the value range of the sample data is [min, max]. The normalized expression is:

[0066]

[0067] Where: k is the serial number of the multivariate information parameter; is the normalized function value of the monitoring data.

[0068] By plotting different physical quantity information in a coordinate system for analysis, key information such as fluctuations and mutations in the original data curve can be reflected. This embodiment takes the test results of a sample with a tangential loading rate of 3.0 MPa / s as an example to perform normalized fusion analysis on the multi-physical field information data during the rock burst disaster. According to the method of formula (1), the physical information such as stress, energy, and acoustic emission are normalized, as shown in the following example: Figure 9 As shown. In addition, the digital image motion analysis software ImageProPlus version 7.0 was used to track the flight trajectory of the ejected fragments and the ejection velocity of the rock fragments was obtained. The mass of the rock burst fragments of four grades was weighed using a high-sensitivity electronic scale, and the particle size distribution curve of the rock burst fragments was obtained using the Weibull distribution function. Figure 9It can be seen that (a) before a rockburst occurs, all four physical quantities exhibit certain precursory characteristics. Points A, B, D, and C represent the precursory characteristic points for stress, energy (total energy, dissipated energy), and acoustic emission, respectively. For example, the stress value at point A reaches the strength of the rock, the strain energy at points B and D exceeds the energy required for rock failure, and the ejection kinetic energy of the fragments increases abnormally. At point C, the acoustic emission event peaks, and tensile or shear fractures cause abnormal changes in the acoustic emission signal. In addition, the occurrence of slab-like splitting of the rock unloading surface and the full development of cracks on the unloading surface can also serve as effective precursor information for rockbursts. (b) Based on the four stages of the rockburst disaster process, the trend of changes in multivariate disaster information can be divided into: initial stage, development stage, pre-disaster stage, and post-disaster stage. (c) During the initial stage, all monitoring information is relatively stable. At this time, the specimen is in a state of stress equilibrium, internal cracks have not yet fully developed, and the average ejection velocity of the fragments on the unloading surface is essentially zero. (d) During the development phase, cracks begin to form within the sample. Although various information changes, they do not reach the critical threshold for rockburst. However, the average ejection velocity of the unloading surface fragments increases significantly. This is mainly due to the sudden increase in the average velocity of the unloading surface caused by the fracture of the rock block surface. That is, when the rock sample fails, part of the elastic strain energy released is converted into kinetic energy, resulting in a higher velocity when the surface rock blocks break. (e) During the disaster-prone phase, various physical field information undergoes sudden changes, and rockburst disasters can occur at any moment in this stage. The average ejection velocity of the unloading surface fragments reaches 4.93 m / s. This indicates that the elastic energy stored in the rock sample is released at the moment of rockburst, a small portion of which is converted into dissipated energy for rock failure, while the majority of the elastic energy is converted into the kinetic energy required for the ejection of fragments. This also indicates that a large portion of the elastic energy released during this stage will be converted into kinetic energy for the ejection of fragments. (f) In the late stage of the disaster, a large number of rock fragments were instantly ejected along the unloading surface, with the average ejection velocity of the fragments on the unloading surface reaching a maximum of 4.97 m / s. This indicates that under the action of vertical stress σ1, the number of split rock slabs caused by tensile stress in the direction perpendicular to the unloading surface increased significantly. Under shear and tensile shear loads, the rock slabs were ejected, forming larger fragments, which caused the average ejection velocity of the fragments on the unloading surface to reach an extreme value. Measuring the mass of the ejected fragments at this time can calculate the maximum initial kinetic energy of the fragments ejected from the parent rock. By the late stage of ejection, large fragments (shards) will completely separate from the parent rock, gradually falling to the ground, and the ejection process ends. The late stage of the disaster also represents a process of secondary rebalancing after the energy balance is broken. Furthermore, further analysis of the ejection velocity of the fragments on the unloading surface shows that the type of rockburst under high tangential loading rates is explosive ejection. (g) The fragments ejected from the unloading surface only have acceleration, and the conversion of acceleration to velocity takes a certain amount of time, which can serve as a warning time. It can be seen that before the rock burst, the velocity value increases sharply and shows an abnormal turning point change. This change feature can be used as a precursor warning information for the rock burst.

[0069] Step 4: Characterize the fusion function of multivariate disaster information changing with time, propose a rockburst multivariate disaster information fusion warning method, and establish a rockburst warning chain.

[0070] Furthermore, the process of establishing a fusion function model includes: modeling the normalized data based on the rational number function fitting method to obtain the fitting functions of stress, acoustic emission, total energy and dissipated energy; performing parameter fitting based on the fitting function to obtain a fitting curve; and obtaining an optimized fusion function model based on the fitting curve.

[0071] Furthermore, the process of determining the key stage division results of the rockburst incubation process includes: based on the first-order derivative and second-order derivative of the fusion function model, analyzing the intersection of the derivative function curve and the time axis to determine the positions of the extreme points and inflection points; determining the dividing point between the pre-disaster period and the post-disaster period based on the positions of the extreme points and the inflection points; obtaining the critical points between the rockburst development period and the pre-disaster period based on the law of change of the function's convexity and concavity; determining the key stage division results of the rockburst incubation process based on the dividing point between the pre-disaster period and the post-disaster period and the critical points between the rockburst development period and the pre-disaster period.

[0072] Furthermore, the process of constructing a rockburst early warning chain includes: based on the key stage division results of the rockburst incubation process, the stress precursor characteristics are obtained through the extreme point analysis of the stress function; based on the extreme point analysis of the acoustic emission function, the acoustic emission precursor characteristics are obtained; based on the inflection point analysis of the total energy function and the dissipated energy function, the energy precursor characteristics are obtained; based on the stress precursor characteristics, the acoustic emission precursor characteristics and the energy precursor characteristics, an early warning mechanism is established; based on the real-time monitoring data of the specimen failure morphology, the destruction precursor characteristics of the unloading surface splitting into slabs and the rock slab bending phenomenon are obtained, and a late warning mechanism is established based on the destruction precursor characteristics; and through the spatiotemporal correlation based on the early warning mechanism and the late warning mechanism, a rockburst early warning chain is constructed.

[0073] Step 4-1: To quantify the evolution characteristics of multivariate information, a mathematical model for function fitting is established to characterize the changing trends of multivariate information. The fitting function should follow the principle of restoring monitoring information and accurately reflect the trend of monitoring data over time. Based on the rational number function, the multivariate information of the entire rockburst disaster process is fitted. The fitting function expression is:

[0074]

[0075] Where: m and n are the orders of the highest order x term in the denominator and numerator respectively; f1(x) is the stress function, f2(x) is the acoustic emission function, f3(x) is the total energy function, and f4(x) is the dissipated energy function; p ij ,q ij is the coefficient of the x term.

[0076] The multivariate information fitting function curve in the rock burst physical simulation test is as follows Figure 10As shown. Figure 10 The correlation coefficient between the fitted curve and the monitored data curve is close to 1, indicating a high degree of fit between the two. This further illustrates that the multivariate information in the rockburst process can be represented in the form of a function, and that applying a functional mathematical model to analyze the evolution of multivariate information is an effective means of monitoring and early warning rockburst disasters.

[0077] Step 4-2: During the rockburst disaster, the multivariate information function changes nonlinearly, such as Figure 10 As shown. Further combined with the sample destruction process, as Figure 9 As shown, and the multivariate information function curve, as Figure 10 As shown in Figure 2, it can be seen that the point where the increase or decrease or concavity of the multivariate information function curve changes may be the dividing point of the rock burst damage stage. In order to clarify the change rate of the normalized multivariate information function to characterize the development process of rock burst damage, Figure 10 The function is first-order and second-order derivatives. When the derivative function curve is Figure 11 As shown, when it intersects the x-axis, the derivative function value is 0, so the intersection point is the stationary point or inflection point.

[0078] contrast Figure 10 and Figure 11 It can be seen that the multivariate information fusion function f i (x) has extreme points and inflection points. The moment corresponding to the maximum point of stress and acoustic emission function is used as the watershed between the pre-disaster period and the post-disaster period of rock burst disaster. The critical point satisfies that the first-order derivative of f1(x) and f2(x) is zero, and the first-order derivative function changes from positive to negative in the neighborhood of zero point. The multivariate information function curve has a situation where the growth rate decreases during the steep increase process, that is, the function changes from a convex function to an up-convex function. The moment when the function undergoes a concave-convex transformation (function stationary point) is the critical point between the rock burst development period and the pre-disaster period. The critical point satisfies that the second-order derivative of f1(x) is zero, and the second-order derivative function changes from positive to negative in the neighborhood of zero point. For a function whose curve change trend is not obvious (a certain section), the method of finding its extreme points and stationary points is no longer applicable. Take f i The time corresponding to the equal division point of the value range of the (x) function within the monitoring time period is taken as the critical value.

[0079] In view of the different weights of the monitoring parameters in the rockburst process, the four critical points obtained by each information function are given different weights for correction, and the rockburst evolution state under different tangential loading rates is obtained as follows: Figure 12 、 Figure 13 、 Figure 14 As shown. Figure 12 、 Figure 13 、 Figure 14It can be seen that (a) the appearance of a yield plateau in the stress-strain curve or the stress value reaching the rock strength is an effective precursor warning signal for rockbursts. The corresponding damage characteristics are splitting failure on the unloading surface of the rock sample and shear failure within the potential rockburst pit. In addition, the peak stress increases with the increase of the tangential loading rate. (b) The number of acoustic emission impacts fluctuates greatly during the frequent period, showing a multi-peak phenomenon, and the signal fluctuation difference during the frequent period shows a characteristic of increasing with the increase of the tangential loading rate. This also indicates that the fractures within the rock are fully incubated and extended, which is an important precursor signal for the impending rockburst. (c) The energy curve shows a turning point, with a more obvious abnormal increase, and the rock then undergoes sudden instability and failure. In addition, the increase in the tangential loading rate increases the energy accumulated within the rock, and the energy within the rock is violently released during the failure, resulting in abnormal changes in the acoustic emission signal. (d) Cracks on the unloading surface fully develop, forming macroscopic fracture zones. Higher tangential loading rates lead to greater abnormal fluctuations in the acoustic emission signal. A sudden increase in microfractures within the rock indicates imminent failure, with the occurrence of rock splitting into slabs and continuous bending of the slabs. (e) The multivariate information function based on normalized fusion not only captures the precursory and sensitivity of multivariate monitoring information but is also more intuitive and convenient for early warning than a multivariate monitoring information data matrix. Under different tangential loading rates, Stages I and II account for a significant portion of the time, representing the energy accumulation phase of rockburst destruction. Stages III and IV, on the other hand, account for a smaller portion of the time, with the rock rapidly failing after reaching its bearing capacity. Therefore, monitoring and early warning of Stages III and IV should be particularly strengthened. (f) This early warning method integrates multiple types of abnormal precursor information for rockburst occurrence. Its greatest advantage lies in its ability to directly, in real time, and continuously obtain comprehensive information on the entire rockburst incubation and development process, enabling timely assessment and prediction of rockburst hazards, thereby providing a useful reference for early warning and prevention of rockburst disasters after rockmass excavation and unloading.

[0080] Based on the above systematic analysis of the abnormal changes of various indicators before rockburst, the multivariate disaster information is correlated and the respective advantages of different information are combined. Starting from the four aspects of rock internal, external, time and space, an effective precursor information early warning chain for rockburst under different tangential loading rate conditions is preliminarily established, such as Figure 15As shown, the information chain includes rock unloading surface fracture precursors, stress precursors, acoustic emission precursors, and energy precursors. Stress values, energy anomalies, and acoustic emission signals in the early warning chain primarily provide temporal warning of rockburst damage, while macroscopic failure surface formation primarily provides spatial warning of rockburst damage. Furthermore, at different tangential loading rates, acoustic emission precursors, energy anomaly precursors, and stress precursors appear relatively early, providing early warning of rock failure; whereas rock cracks appear relatively late. Therefore, acoustic emission precursors, energy anomaly precursors, and stress precursors can serve as early warnings in the rockburst precursor information chain, prioritizing rock failure in time. Rock crack propagation, on the other hand, can serve as later warnings, providing spatial warnings of potential rock failure locations. In summary, the rockburst precursor information chain provides both temporal and spatial warnings of rock failure. Early warnings enable early response to rock failure, while late warnings allow for targeted responses to potential rock failure locations.

[0081] The present invention provides a rockburst early warning method based on multivariate disaster information fusion. Based on a true triaxial rockburst test machine and equipped with a multivariate disaster information monitoring system, a true triaxial test of "radial rapid unloading-three-way five-surface force-tangential continuous loading" was carried out to study the multivariate information such as the entire process of rockburst disaster, destruction morphological characteristics, peak intensity, acoustic emission characteristics, fragment ejection and energy evolution, clarify the multivariate disaster information correlation effect based on normalized fusion analysis, and establish a rockburst multivariate disaster information fusion early warning method. The multivariate information function based on normalized fusion established by the present invention not only includes the precursory and sensitivity of multivariate monitoring information, but is also more intuitive and more convenient for early warning than the multivariate monitoring information data matrix. In addition, the rockburst precursor information chain can simultaneously warn of rock damage in time and space. Early warning can enable people to respond to rock damage early, while late warning can enable people to respond to potential rock damage locations. At the same time, this early warning method integrates multiple types of abnormal precursor information of rockburst. Its greatest advantage is that it can directly, in real time and continuously obtain comprehensive information on the entire process of rockburst incubation and development, and timely evaluate and predict the danger of rockburst, in order to provide useful reference for early warning and prevention of rockburst disasters after rock excavation and unloading.

[0082] The above are merely preferred embodiments of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

Claims

1. A rockburst early warning method based on multi-dimensional disaster information fusion, characterized in that: The following steps are involved: The standard rock specimens were placed in a true triaxial test system, and the rockburst incubation process was simulated through radial rapid unloading and tangential continuous loading, while basic test data was collected simultaneously. Normalizing the experimental basic data to obtain a normalized data set with unified dimensions; Establishing a fusion function model based on the normalized data set, and determining the key stage division results of the rockburst incubation process by analyzing the time series distribution characteristics of the extreme points and inflection points of the fusion function model; Based on the key stage division results of the rockburst incubation process, a rockburst early warning chain is constructed from the internal state, external performance, time series and spatial distribution of the rock to achieve early warning of rockburst disasters.

2. The rockburst early warning method based on multi-dimensional disaster information fusion according to claim 1 is characterized in that: The process of collecting basic test data includes: Based on the same piece of granite, rectangular prisms were prepared to obtain standard rock samples; A fully digital servo controller is used to control the independent loading and unloading in the vertical and horizontal directions of the true triaxial test system, and the test is carried out according to the radial rapid unloading and tangential continuous loading path; at the same time, the acquisition equipment synchronously collects multi-dimensional disaster information and transmits the multi-dimensional disaster information in real time to the computer for storage and preliminary processing to obtain basic test data.

3. The rockburst early warning method based on multi-dimensional disaster information fusion according to claim 1 is characterized in that: The experimental basic data are processed based on the deviation standardization method to obtain a normalized data set with unified dimension.

4. The rockburst early warning method based on multi-dimensional disaster information fusion according to claim 3 is characterized in that: The process of establishing a fusion function model includes: The normalized data were modeled based on the rational number function fitting method to obtain the fitting functions of stress, acoustic emission, total energy and dissipated energy; Performing parameter fitting on the fitting function to obtain a fitting curve; An optimized fusion function model is obtained based on the fitting curve.

5. The rockburst early warning method based on multi-dimensional disaster information fusion according to claim 4 is characterized in that: The fusion function model is: Where, f i (x) is the fusion function model, m and n are the orders of the highest order x term in the denominator and numerator respectively; f1(x) is the stress function, f2(x) is the acoustic emission function, f3(x) is the total energy function, and f4(x) is the dissipated energy function; p ij ,q ij is the coefficient of the x term.

6. The rockburst early warning method based on multi-dimensional disaster information fusion according to claim 1 is characterized in that: The process of determining the key stages of the rockburst incubation process includes: Based on the first-order derivative and second-order derivative of the fusion function model, the intersection of the derivative function curve and the time axis is analyzed to determine the location of the extreme points and inflection points; Determine the dividing point between the pre-disaster period and the post-disaster period based on the positions of the extreme points and the inflection points; Based on the changing law of the function's concavity and convexity, the critical points between the rockburst development period and the impending disaster period are obtained. The key stage division results of the rockburst incubation process are determined based on the dividing point between the pre-disaster period and the post-disaster period and the critical point between the rockburst development period and the pre-disaster period.

7. The rockburst early warning method based on multi-dimensional disaster information fusion according to claim 6 is characterized in that: The process of building a rockburst early warning chain includes: Based on the key stage division results of the rockburst incubation process, stress precursor characteristics are obtained through extreme point analysis of the stress function; Based on the extreme point analysis of the acoustic emission function, the acoustic emission precursor characteristics are obtained; Based on the inflection point analysis of the total energy function and the dissipated energy function, the energy precursor characteristics are obtained; Establishing an early warning mechanism based on the stress precursor characteristics, acoustic emission precursor characteristics and energy precursor characteristics; Based on the real-time monitoring data of the specimen failure morphology, the failure precursor characteristics of the unloading surface splitting into slabs and the bending of the rock slab are obtained, and a late warning mechanism is established based on the failure precursor characteristics; Through the temporal and spatial association of the early warning mechanism and the late warning mechanism, a rock burst warning chain is constructed.

Images