A method for distinguishing acoustic emission-fracturing pressure time sequence in rock mass hydraulic fracturing stage

Abstract

The application discloses a rock mass hydraulic fracturing stage acoustic emission-fracturing pressure timing discrimination method, synchronously collects rock mass acoustic emission signals and fracturing pressure data, respectively calculates acoustic emission energy release rates and fracturing pressure change rates, adopts a unified sliding window to complete cumulative smoothing, constructs a cumulative acoustic emission energy release rate curve and a cumulative fracturing pressure change rate curve, identifies characteristic inflection points of the two curves and calculates the inflection point time difference, combines the curve evolution law and the time difference characteristics, and automatically divides the hydraulic fracturing process into four stages of compaction, cracking, expansion and penetration, and stable pressure diffusion, increases the result stability check, outputs the discrimination data, and can feed back the result to a fracturing control system to realize process regulation and control. The method realizes automatic and standardized discrimination of the hydraulic fracturing stage, can guide the optimization of field fracturing process parameters, reduces the risk of water inrush, rock burst and other dynamic disasters, and is suitable for various deep resource hydraulic mining scenes.

Description

Technical Field

[0001] This invention belongs to the field of mineral mining technology, and in particular relates to a method for determining the timing of acoustic emission and fracturing pressure during the hydraulic fracturing stage of rock mass. Background Technology

[0002] With the increasing depletion of shallow mineral resources, deep resource extraction has become a crucial means of ensuring energy supply. Deep fluidized bed mining, a cutting-edge technology for the efficient development of deep solid mineral resources, involves injecting fluids (water or fracturing fluid with additives) under high pressure into deep rock masses. This allows the rock fracture network to fully develop and form a synergistic transport channel with the fluidized medium, thereby achieving fluidized ore extraction. During deep fluidized bed mining, the continuous injection of high-pressure fluid induces hydraulic fracturing within the rock mass. The initiation, expansion, and connection of fractures directly determine the seepage path of the fluidized medium and the mining efficiency, while also affecting the stability of the surrounding rock and operational safety. Therefore, accurately identifying the evolution stages of hydraulic fracturing in the rock mass (compaction, initiation, expansion and connection, and pressure stabilization and diffusion) is of significant engineering importance for optimizing fracturing process parameters and preventing dynamic disasters such as water inrush and roof collapse. Currently, acoustic emission (AE) technology is widely used for monitoring and stage identification of rock fracture processes due to its ability to capture elastic wave signals released from micro-fractures within the rock mass in real time. Existing AE stage identification methods mainly fall into two categories: The first category is stage division methods based on single AE parameter thresholds or statistical characteristics. For example, using the inflection point of the cumulative curve of AE ring count, energy release rate, or b-value, the conventional loading process of rock (uniaxial or triaxial compression) can be divided into compaction, elastic, damage, and failure stages; or the fracture penetration time can be determined by the spatial distribution density change of AE events. These methods are relatively effective in conventional mechanical tests, but in the high-pressure hydraulic-driven environment of deep fluidized bed mining, the superimposed interference of fluid injection noise, mechanical vibration noise, and particle collision noise can easily cause fluctuations and abnormal jumps in single AE parameters, posing a challenge to the stability and reliability of stage identification results. The second category is intelligent monitoring and early warning systems based on multimodal sensors and deep learning. For example, by deploying multi-channel acoustic emission / microseismic sensor arrays, combined with environmental noise modeling and convolutional neural network denoising, rock fracturing events can be identified and seismic source located, leading to graded early warning. Such systems can handle complex noise environments, but they are essentially system-level solutions relying on a large amount of hardware, complex algorithm models, and training samples. This results in high deployment costs, high computational resource requirements, and limited real-time performance, hindering their rapid application in general fluidized bed mining faces. Furthermore, in the current field of hydraulic fracturing, the identification of fracturing stages is often subjectively estimated by engineers through visual interpretation of the morphological characteristics of the fracturing pressure-time curve (such as pressure peaks and descent inflection points). While this method can reflect the fracturing process to some extent, a quantitative time-series correlation model between fracturing pressure response and acoustic emission energy release has not yet been established. This makes it difficult to objectively reflect the physical causal relationship between "pressure-driven—fracture propagation—energy release," and the accuracy of its discrimination needs further improvement regarding the nonlinear and dynamic characteristics of the interaction between fracturing fluid and rock mass in deep fluidized bed mining.

[0003] In summary, the existing technologies have the following main shortcomings and challenges in the identification of hydraulic fracturing stages in deep resource mining: (1) Insufficient robustness of single-parameter identification: They rely heavily on threshold or inflection point identification of a single acoustic emission parameter (ring count, energy or b-value). In the complex noise environment of deep fluidized mining, the stage transition characteristics are easily interfered with, and the stability of the identification results needs to be enhanced; (2) Limited universality of system-level solutions: Monitoring and early warning systems based on multimodal arrays and deep learning have strong hardware dependence and high algorithm complexity, making them unsuitable for universal on-site rapid response. When using rapid identification methods, their applicability faces certain limitations; (3) lack of pressure-acoustic emission time-series correlation mechanism: existing methods have failed to effectively establish time-series correlation criteria between fracturing pressure change rate and acoustic emission energy release rate, making it difficult to quantitatively characterize the physical law of "pressure response ahead and energy release lag" under hydraulic drive, and the mechanical basis for stage division is still insufficient; (4) low degree of standardization of manual interpretation: manual interpretation methods that rely on experience lack unified quantitative standards, and there are still certain technical bottlenecks in realizing automated and quantitative identification of fracturing stages.

[0004] To address the shortcomings of existing technologies, there is an urgent need to propose a rock mass hydraulic fracturing stage discrimination method that is applicable to deep resource mining scenarios, relies solely on the dual parameters of acoustic emission and fracturing pressure, and has a simple and clear algorithm, in order to overcome the limitations of existing methods in terms of stability, universality, and mechanical basis. Summary of the Invention

[0005] The purpose of this invention is to provide a method for determining the timing of acoustic emission and fracturing pressure during the hydraulic fracturing stage of rock mass, thereby solving the technical problems in the prior art, such as weak anti-interference ability of single parameters, poor universality of complex systems, lack of quantitative timing correlation between pressure and acoustic emission signals, and lack of standardized manual determination.

[0006] The technical solution adopted by this invention to solve its technical problem is: a method for determining the timing of acoustic emission and fracturing pressure during the hydraulic fracturing stage of rock mass, comprising the following steps:

[0007] S1. Multi-source data synchronous acquisition: Acoustic emission sensors and fracturing pressure sensors are simultaneously deployed in the hydraulic fracturing operation area of ​​deep resource mining to collect rock mass acoustic emission signals and fracturing pipeline pressure data, and to synchronize the two types of data in time.

[0008] S2. Calculation and cumulative smoothing of acoustic emission energy release rate: Divide the acoustic emission signal into segments according to a fixed time window length Δt0, calculate the acoustic emission energy and acoustic emission energy release rate SER(i) in each time window, and use a fixed length sliding window to perform cumulative smoothing on SER(i) to construct the cumulative acoustic emission energy release rate curve CSER(t).

[0009] S3. Calculation and cumulative smoothing of fracturing pressure change rate: Divide the fracturing pressure data into segments according to the same time window length Δt0 as in step two, calculate the fracturing pressure change rate PVR (i) in each time window, and use a sliding window with the same parameters as in step two to perform cumulative smoothing on PVR (i) to construct the cumulative fracturing pressure change rate curve CPVR (t).

[0010] S4. Characteristic Inflection Point Identification of Two-Parameter Cumulative Curve: Identifying the peak inflection point time t of the CPVR(t) curve. P Simultaneously identify the sudden inflection point t of the CSER(t) curve. E ;

[0011] S5. Energy-Pressure Response Time Difference Calculation: Based on the formula Calculate the response time difference between the inflection point of the sudden increase in acoustic emission energy and the inflection point of the peak fracturing pressure. ;

[0012] S6. Hydraulic fracture stage identification: Combining the morphological evolution characteristics, inflection point order, and response time difference of the CPVR(t) and CSER(t) curves. The hydraulic fracturing process of rock mass is classified into corresponding stages in the compaction stage, fracturing initiation stage, propagation and penetration stage, and pressure stabilization and diffusion stage.

[0013] S7. Result Verification, Output, and Process Feedback: Set continuous sliding window stability verification rules. After successful verification, the output includes the current stage, stage time, duration, CSER(t), CPVR(t), and response time difference. The judgment results are included; at the same time, the judgment results are fed back to the hydraulic fracturing process control system in real time to adjust the injection rate, pump pressure limit and / or pump shutdown timing.

[0014] In the acoustic emission-fracturing pressure timing discrimination method for the hydraulic fracturing stage of rock mass according to the present invention, in step s2, the formula for calculating the acoustic emission energy within the i-th time window is as follows:

[0015]

[0016] In the formula, N i A is the ringing count of acoustic emission events within the i-th time window; j Given the peak amplitude of the j-th acoustic emission event, calculate the acoustic emission energy release rate SER(i) for the i-th time window:

[0017] Calculate the acoustic emission energy release rate SER(i) for the i-th time window:

[0018]

[0019] Given a sliding window length of W, the formula for the cumulative acoustic emission energy release rate corresponding to the k-th sliding window is:

[0020]

[0021] In the acoustic emission-fracturing pressure time sequence discrimination method for the hydraulic fracturing stage of rock mass according to the present invention, the formula for calculating the fracturing pressure change rate within the i-th time window in step s3 is as follows:

[0022]

[0023] In the formula, Let fracturing pressure be the value at the end of the i-th time window; For the first The fracturing pressure value at the end of each time window, where Δt0 is the fixed time window length;

[0024] The formula for calculating the cumulative fracturing pressure change rate corresponding to the kth sliding window is:

[0025] .

[0026] The present invention provides a method for determining the acoustic emission-fracturing pressure timing during the hydraulic fracturing stage of rock mass, wherein the peak inflection point of the CPVR(t) curve in s4 is... satisfy:

[0027]

[0028] Right now The moment when the first derivative of the curve changes from positive to negative and the second derivative becomes negative indicates that the driving force of the fracturing pressure growth changes from dominant to attenuated, corresponding to the critical state of the rock mass that is about to or has already begun hydraulic fracturing.

[0029] The sudden inflection point of the CSER(t) curve satisfy:

[0030]

[0031] The current slope of the CSER(t) curve is greater than the average slope of the previous baseline. The value is multiplied by a factor of m, and this state persists for at least m sliding windows, where: Let be the average slope of the baseline segment of the CSER(t) curve before time t; The threshold value is the sudden increase factor, and m is the preset duration window number.

[0032] The present invention provides a method for determining the acoustic emission-fracturing pressure timing sequence during the hydraulic fracturing stage of rock mass. The specific stage determination conditions in step S6 are as follows:

[0033] During the compaction phase: CPVR(t) continues to increase monotonically, CSER(t) is below the preset baseline threshold S1 and remains stable, and no valid t is identified. P With t E Inflection point, or Exceeds the valid discrimination range;

[0034] Initiation stage: CPVR(t) passes through the peak inflection point t P Afterwards, it turns downward, and CSER(t) shows a sudden increase at the inflection point t. E And satisfy 0 < Δt < Δt theory ·(1+γ), where Δt theory γ is the theoretical time difference, and γ is the allowable deviation coefficient;

[0035] During the expansion and connection phase: CSER(t) remains in the high range above the threshold S2 or experiences a second surge, CPVR(t) continues to decrease or fluctuates slightly, and Δt tends to stabilize;

[0036] During the voltage stabilization and diffusion stage: CSER(t) changes from a high level to a decrease, and CPVR(t) converges to the stabilizing voltage P. steady Furthermore, if the fluctuation amplitude is less than the preset threshold, Δt will no longer be used as the main criterion.

[0037] The present invention provides a method for determining the timing of acoustic emission-fracturing pressure during the hydraulic fracturing stage of rock mass. The stability verification rule is as follows: when the stage determination results of 3-5 consecutive sliding windows are consistent, the verification is deemed successful and the determination result is officially output.

[0038] The present invention provides a method for determining the timing of acoustic emission-fracturing pressure during the hydraulic fracturing stage of rock mass. In step s2, the acoustic emission ringing count rate or the acoustic emission event amplitude change rate is used to replace the acoustic emission energy release rate to complete the construction of the cumulative curve and the identification of the inflection point.

[0039] The present invention provides a time-series discrimination method for acoustic emission-fracturing pressure during the hydraulic fracturing stage of rock mass. Based on step s3, the method simultaneously collects fracturing injection rate data, calculates the injection rate change rate and constructs a cumulative injection rate change rate curve, and combines CSER(t) and CPVR(t) to form a three-parameter time-series correlation discrimination system.

[0040] The beneficial effects of this invention are:

[0041] 1. Strong anti-interference ability and high robustness of judgment: The dual-parameter verification of acoustic emission and fracturing pressure is adopted to cancel out the abnormal fluctuations of parameters caused by on-site noise, thus solving the problem of easy misjudgment by single parameter judgment.

[0042] 2. Sufficient physical basis and standardized judgment: Based on the physical law of "pressure-driven-crack initiation-energy release", a time-series time difference criterion is established to get rid of manual experience-based judgment and realize quantitative and standardized judgment throughout the process.

[0043] 3. Simple algorithm and wide engineering applicability: It only includes basic algorithms such as data operation, smoothing and inflection point identification. It does not require multi-channel arrays or deep learning models. It has low hardware cost and low computational load. It can be directly embedded into existing fracturing monitoring systems, which is convenient for large-scale promotion.

[0044] 4. Stable output results and strong production guidance: Multi-window stability verification is added to avoid instantaneous noise causing result jumps; the judgment results can be directly linked to the fracturing control system to optimize process parameters, improve rock mass permeability efficiency, and effectively avoid dynamic disasters such as water inrush, rock bursts, and roof falls.

[0045] 5. High scalability: Supports various transformation schemes such as parameter replacement, multi-parameter expansion, and inflection point algorithm replacement, and can be adapted to various application scenarios such as laboratory tests and different types of deep mining faces. Attached Figure Description

[0046] The present invention will be further described below with reference to the accompanying drawings and embodiments. In the accompanying drawings:

[0047] Figure 1 This is a flowchart of an embodiment of the present invention; Detailed Implementation

[0048] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0049] like Figure 1 As shown, a method for determining the acoustic emission-fracturing pressure timing during the hydraulic fracturing stage of rock mass is presented:

[0050] Step 1: Synchronous Acquisition of Multi-Source Data

[0051] In deep resource extraction hydraulic fracturing operation areas, acoustic emission sensors and fracturing pressure sensors are deployed simultaneously to construct a dual-parameter synchronous acquisition system. Acoustic emission signal acquisition: High-sensitivity acoustic emission sensors are deployed on the rock surface surrounding the fracturing hole or on the borehole inner wall. These sensors are then connected to the data acquisition system via a preamplifier and anti-aliasing filter, with a sampling frequency of no less than 1 MHz to ensure the complete waveform of micro-fracture events is captured. Fracturing pressure data acquisition: High-precision fluid pressure sensors are deployed near the borehole opening in the fracturing pipeline to accurately record the dynamic response process of the fracturing pressure. Time synchronization: The acoustic emission data acquisition system and the pressure data acquisition system are synchronized via GPS / BeiDou timing or hardware triggering to ensure that both types of data have a unified time reference.

[0052] Step 2: Calculation and cumulative smoothing of acoustic emission energy release rate

[0053] Energy calculation and adaptive cumulative smoothing are performed on the synchronously acquired acoustic emission signals to construct the cumulative acoustic emission energy release rate curve.

[0054] (1) Calculation of energy release rate

[0055] The acoustic emission signal is divided into segments according to a fixed time window length Δt0, and the acoustic emission energy E(i) in the i-th time window is calculated:

[0056]

[0057] In the formula, N i A is the ringing count of acoustic emission events within the i-th time window; j Let be the peak amplitude of the j-th acoustic emission event. Calculate the acoustic emission energy release rate SER(i) for the i-th time window:

[0058] Calculate the acoustic emission energy release rate SER(i) for the i-th time window:

[0059]

[0060] (2) Adaptive sliding window cumulative smoothing

[0061] A fixed-length sliding window is used to cumulatively smooth SER(i). The sliding window length is set to W. The cumulative acoustic emission energy release rate CSER(k) within the k-th sliding window is:

[0062]

[0063] By traversing the entire fracturing process through a sliding window, the cumulative acoustic emission energy release rate time series CSER(t) is obtained. This cumulative smoothing process can effectively filter out the instantaneous impulse noise of the acoustic emission signal and highlight the overall trend evolution characteristics of energy release.

[0064] Step 3: Calculation and cumulative smoothing of fracturing pressure change rate

[0065] The rate of change and cumulative smoothing processing of the synchronously acquired fracturing pressure data are performed to construct the cumulative fracturing pressure rate of change curve.

[0066] (1) Calculation of pressure change rate

[0067] The fracturing pressure data is segmented according to the same time window length Δt0 as in step two, and the fracturing pressure change rate PVR(i) within the i-th time window is calculated:

[0068]

[0069] In the formula, Let fracturing pressure be the value at the end of the i-th time window; For the first The fracturing pressure value at the end of the time window.

[0070] (2) Cumulative smoothing

[0071] Using the exact same sliding window length W and step size as in step two, PVR(i) is cumulatively smoothed. The cumulative fracturing pressure change rate within the k-th sliding window... for:

[0072]

[0073] The cumulative fracturing pressure change rate time series CPVR(t) is obtained by iterating through a sliding window. This adaptive cumulative smoothing process can smooth out the sampling jitter of the pressure sensor while preserving the statistical consistency between the pressure response and energy release sequences, making the two cumulative curves comparable.

[0074] Step 4: Identification of Inflection Points in Two-Parameter Cumulative Curves

[0075] The two cumulative curves CSER(t) and CPVR(t) obtained in steps two and three are used to identify characteristic inflection points and determine the time points with physical significance.

[0076] (1) Peak inflection point identification

[0077] The curve reflects the cumulative effect of the rate of change of fracturing pressure. During the compaction stage, as fracturing fluid is continuously injected, The overall trend is upward; when internal fractures in the rock mass begin to initiate, fracturing fluid is lost, and the rate of pressure increase slows down. The change from rising to falling. Definition Peak inflection point of the curve for:

[0078]

[0079] Right now The moment when the first derivative of the curve changes from positive to negative and the second derivative becomes negative indicates that the driving force of the fracturing pressure growth changes from dominant to attenuated, corresponding to the critical state of the rock mass that is about to or has already begun hydraulic fracturing.

[0080] (2) Sudden Inflection Point Identification

[0081] The curve reflects the cumulative trend of acoustic emission energy release. During compaction and the early stages of crack initiation, acoustic emission activity is weak because the cracks have not yet fully expanded. Maintain a low and stable level; when the crack enters the rapid propagation stage, micro-fracture events increase dramatically. The trend shifted from steady to rapid growth. (Definition) The moment t_E at the sudden inflection point of the curve is:

[0082]

[0083] In the formula, Let be the average slope of the baseline segment of the CSER(t) curve before time t; is the threshold for the sudden increase factor; m is the number of sustained windows. This inflection point characterizes the transition of acoustic emission energy release from the background level to a significantly active state, corresponding to the rapid expansion of the fracture network.

[0084] Step 5: Calculation of Energy-Pressure Response Time Difference

[0085] Calculate the response time Δt between the inflection point of the acoustic emission energy surge and the inflection point of the peak fracturing pressure:

[0086]

[0087] The physical significance of this time difference lies in the fact that the change in fracturing pressure precedes the significant release of acoustic emission energy, reflecting the causal sequence of "pressure-driven fracturing initiation, followed by the release of elastic energy and acoustic emission after fracturing expansion." In the hydraulic fracturing process of deep resource extraction, this time difference usually satisfies Δt > 0, and its magnitude is closely related to rock mass integrity, fracturing fluid viscosity, and injection rate.

[0088] Step Six: Identification of Hydraulic Fracturing Stage

[0089] Based on the morphological characteristics, inflection point order, and response time difference Δt of the two cumulative curves CPVR(t) and CSER(t), the hydraulic fracturing process is identified into the following four stages:

[0090] Stage I - Compaction Stage

[0091] Judgment criteria: CPVR(t) continuously increases monotonically; CSER(t) remains stable at a low level, CSER(t) < ( (This is the baseline threshold for energy release rate, calibrated through indoor testing); no effective [measure / method] has yet been found. and Inflection point, or Δt exceeding the effective discrimination range. Physical meaning: The fracturing fluid mainly compresses the pores of the rock matrix and existing closed fractures, without generating new hydraulic fractures, and acoustic emission activity is at the background level.

[0092] Stage II - Cracking Initiation Stage

[0093] Judgment condition: CPVR(t) shows a peak inflection point. Then it begins to decline; CSER(t) shows a sudden inflection point. The response time difference satisfies 0 < Δt < Δt theory ·(1+γ). Where γ is the allowable deviation coefficient. This discrimination condition replaces the artificially calibrated fixed threshold Δt with the theoretically calculated value based on the rock mass physical parameters, so that the stage discrimination has a clear mechanical causal basis.

[0094] Physical significance: The fracturing fluid pressure overcomes the rock mass initiation pressure, and hydraulic fractures begin to emerge; since fracture propagation requires time to accumulate, the significant release of acoustic emission energy lags behind the pressure peak response. The time difference Δt quantitatively characterizes this lag effect, and its time difference range is determined by the physical properties of the rock mass rather than being artificially set.

[0095] Phase III - Extension and Connection Phase

[0096] Judgment criteria: CSER(t) remains above a high level or experiences a second surge; CPVR(t) continues to decline or enters a state of slight fluctuation (absolute value less than the threshold δ). P The response time difference Δt tends to stabilize or no longer increases significantly. Physical significance: The hydraulic fracture network expands rapidly and interconnects with each other, fracturing fluid leakage channels are formed, resulting in a continuous decrease in pipeline pressure; microfracture events occur intensively inside the rock mass, and acoustic emission energy remains at a high level.

[0097] Stage IV - Stabilized Diffusion Stage

[0098] Judgment criteria: CSER(t) shows an inflection point from a high level to a decrease (energy release rate decays); CPVR(t) tends to a stable value P. steady (Stable voltage value), fluctuation range is less than the threshold ε P The response time difference Δt is no longer used as the main control indicator for stage discrimination.

[0099] Physical significance: The fracture system is basically developed, the fracturing fluid achieves steady-state diffusion and filling in the formed fracture network, the rock mass fracturing activity weakens, and the system enters a relatively stable fracturing maintenance state.

[0100] Step 7: Output of Judgment Results and Process Feedback

[0101] The stage discrimination results are output in time series form, including: current stage number, stage name, stage start time, duration, current CSER(t) and CPVR(t) values ​​and response time difference Δt.

[0102] Optionally, the judgment results can be fed back to the hydraulic fracturing process control system for deep resource extraction in real time as a decision-making reference for adjusting the injection rate, pump pressure limit, or pump shutdown timing. For example, when the judgment results change from Stage II to Stage III, the injection rate can be appropriately reduced to control excessive fracture propagation; when the judgment results enter Stage IV and remain stable, they can be used as a criterion for stopping the pump or switching the grouting process.

[0103] (1) Effects of the two-parameter temporal correlation discrimination mechanism

[0104] By simultaneously introducing the acoustic emission energy release rate and the fracturing pressure change rate, and establishing the time-series correlation of the inflection points of their cumulative curves ( This effectively overcomes the shortcomings of single-parameter discrimination, which is susceptible to interference in the high-noise environment of deep fluidized bed mining. When the acoustic emission signal exhibits instantaneous abnormal jumps due to mechanical vibration or fluid noise, the fracturing pressure parameter can provide independent trend constraints; conversely, when the pressure sensor generates spurious fluctuations due to pipeline fluctuations, the acoustic emission energy parameter can provide evidence of physical fracturing events. The mutual verification of the two parameters significantly improves the robustness and reliability of the stage discrimination results.

[0105] (2) Effects of adaptive cumulative smoothing

[0106] Instead of manually setting a fixed window length, an adaptive cumulative smoothing method is used to perform smoothing of energy release rate and pressure change rate based on signal autocorrelation characteristics. This adaptive mechanism matches the window length with the statistical structure of the signal itself, effectively suppressing sensor sampling noise and environmental impulse noise while preserving the trend evolution characteristics of the signal. Shifting the focus of discrimination from "instantaneous state" to "trend evolution" is more in line with the physical nature of hydraulic fracturing of rock mass as a gradual cumulative process, and eliminates the need for manual adjustment of window parameters.

[0107] (3) Effects of the physical coupling time difference criterion

[0108] The theoretical time difference expression Δt derived using stress wave propagation theory theoryThis method transforms the physical law of "preceding pressure response and delayed acoustic emission energy release" into mathematical indicators related to rock mass physical parameters (P-wave velocity, fracture toughness, and elastic modulus). A verification mechanism based on the theoretical-to-measured time difference deviation rate ensures the reliability of the judgment basis. This criterion not only provides a clear mechanical basis for stage division (pressure-driven fracturing initiation, fracturing initiation leading to energy release), but also represents a leap from qualitative manual interpretation to judgment based on a physical model, significantly improving the standardization and repeatability of the judgment results.

[0109] (4) The effects of the four-stage division system

[0110] The hydraulic fracturing process for deep resource extraction is divided into four consecutive stages: compaction, fracturing initiation, propagation and penetration, and pressure stabilization and diffusion, covering the complete physical process from fluid injection to fracture stabilization. This classification system highly corresponds to the field fracturing process (pressure boosting, pressure stabilization, pressure depressurization, and pump shutdown), and can directly serve the real-time optimization of fracturing parameters. It helps to improve the fracture permeability enhancement efficiency of fluidized bed extraction and reduce the risk of disasters such as water inrush and rock bursts caused by uncontrolled fracturing.

[0111] (5) The effect of algorithm simplicity

[0112] This invention maintains a simple algorithm (involving only data synchronization acquisition, differential calculation, adaptive sliding window accumulation, and physical coupling inflection point discrimination) while eliminating reliance on a large number of manually calibrated parameters through an adaptive threshold mechanism. It has low computational complexity, requiring no multi-channel sensor arrays, deep learning models, or large-scale training samples. This method can be easily embedded into the software modules of existing hydraulic fracturing monitoring systems, achieving real-time stage discrimination in the field at a low cost. It also possesses field self-calibration capabilities, eliminating the need for re-calibrating indoor tests for each new construction site, and has significant engineering application value.

[0113] Implementation Method 1: Model Calibration and Verification Based on Indoor True Triaxial Hydraulic Fracturing Tests

[0114] Hydraulic fracturing simulation tests were conducted on coal and rock samples under laboratory conditions using a true triaxial loading system. Acoustic emission signals and fracturing pressure data were simultaneously acquired during the test. CSER(t), CPVR(t), and Δt were calculated according to steps two through six. Threshold parameters were calibrated by combining data on fracture development obtained from high-speed imaging or CT scans. , , , , And the burst factor λ. This implementation method is mainly used for the initial calibration of method parameters and mechanism verification.

[0115] Implementation Method 2: Real-time Judgment Based on On-site Hydraulic Fracturing Monitoring

[0116] In deep resource extraction sites, acoustic emission sensors are deployed in the surrounding rock of the roadway or inside the borehole, while pressure sensors are connected to the fracturing pipeline. An industrial control computer, through on-site data acquisition, executes the calculation process from steps two to seven in real time. The stage judgment result is updated after each sliding window of data processing is completed, and the current fracturing stage is displayed in real time via a human-machine interface. This implementation method directly supports the optimization of on-site fracturing processes.

[0117] Modification Scheme 1: Substitution of Acoustic Emission Parameters

[0118] In step two, the acoustic emission energy E(i) can be replaced by the acoustic emission ring count N(i) or the sum of squares of the amplitudes of acoustic emission events, and the corresponding energy release rate SER(i) can be replaced by the ring count rate or the amplitude change rate. The cumulative smoothing method remains unchanged, and the cumulative curve can still be constructed and the inflection point identified.

[0119] Deformation Scheme Two: Expansion of Fracturing Parameters

[0120] In step three, in addition to fracturing pressure P(t), injection rate Q(t) or flow rate change rate can be introduced simultaneously to construct a cumulative injection rate change rate curve, which forms a three-parameter time series correlation with CSER(t), further improving the redundancy and reliability of stage discrimination.

[0121] Modification Solution 3: Replacement of Inflection Point Recognition Algorithm

[0122] In step four, in addition to using the sign change of the first derivative and the second derivative criterion, the inflection point identification can also use the cumulative sum control chart (CUSUM) algorithm, the Bayesian change point detection algorithm, or the wavelet transform modulus maxima method. All of these are equivalent variations of the "cumulative curve inflection point identification" technical concept of this invention.

[0123] Modification Scheme 4: Adjustment of Stage Division

[0124] Depending on the specific engineering requirements, the four stages can be divided into three stages (merging the expansion and penetration and the pressure stabilization diffusion into the "expansion and pressure stabilization stage") or five stages (adding a "pre-compression stage" before the crack initiation stage). Adjusting the number of stages and the discrimination conditions does not change the core technical idea of ​​the present invention, "time sequence discrimination of inflection point of dual-parameter cumulative curve".

[0125] Modification Scheme 5: Closed-Loop Control Integration

[0126] The stage determination results from step seven are linked with the fracturing pump control system to construct a closed loop of "monitoring-determination-control". For example, when the determination enters stage III, the pump pressure is automatically reduced; when the determination enters stage IV and remains stable, the pump stop command is automatically executed, realizing intelligent stage adaptive control of the hydraulic fracturing process.

[0127] Compared with existing technologies for monitoring and stage discrimination of hydraulic fracturing in deep resource mining, the key difference of this invention lies in its breakthrough from the traditional paradigm of single-parameter threshold discrimination and multi-modal system-level monitoring. It establishes a simplified and quantitative stage discrimination mechanism based on the time-series correlation of the cumulative curves of acoustic emission energy release rate and fracturing pressure change rate. Specifically, the key differences and technical effects of this invention are as follows:

[0128] Key difference 1: Simultaneous acquisition of two parameters and construction of adaptive cumulative curves

[0129] Existing technologies often rely on a single acoustic emission parameter (ring count, energy, or b-value) for stage discrimination, or on multi-channel sensor arrays and complex algorithms to construct monitoring systems. Furthermore, cumulative smoothing typically employs a manually set fixed window length. This invention simultaneously acquires only two types of parameters: acoustic emission signals and fracturing pressure data, and calculates the energy release rate SER(i) and pressure change rate PVR(i) separately. It innovatively employs a dynamic sliding window based on signal autocorrelation characteristics for adaptive cumulative smoothing. The window length W is determined by the first zero-crossing point of the autocorrelation function of the acoustic emission energy release rate sequence, replacing the traditional manually fixed value. This difference enables the cumulative curve construction process to be data-driven and adaptive, eliminating the need for manual adjustment of window parameters, significantly reducing the complexity of data acquisition and processing, and avoiding the costs and algorithmic dependencies associated with multimodal systems.

[0130] Key difference two: Temporal correlation criterion for two-parameter cumulative curves based on physical coupling model

[0131] Existing technologies for identifying fracturing stages largely rely on manual visual interpretation of fracturing pressure curves or identification of independent inflection points on single-parameter acoustic emission curves. These methods fail to establish a quantitative temporal correlation between the two types of parameters, and the time difference is merely a simple calculation. This invention innovatively establishes a theoretical time difference expression based on stress wave propagation theory, physically linking the time difference between fracturing pressure response and acoustic emission energy release with physical properties such as rock mass longitudinal wave velocity and fracture toughness. Simultaneously, a verification mechanism for the theoretical-measured time difference deviation rate is proposed to ensure the reliability of the judgment basis. This distinction provides a clear mechanical causal basis and physical model support for stage division, achieving a leap from subjective experience-based interpretation to quantitative judgment based on a physical model.

[0132] Key difference three: A four-stage collaborative discrimination matrix based on physical constraint response time difference and curve evolution trend.

[0133] Existing technologies for stage division often employ single threshold triggers or independent parameter thresholds, and the time difference threshold relies on manual calibration. This invention establishes a four-stage collaborative discrimination matrix with theoretical time difference and its allowable deviation range as the main control index, and the subsequent evolution trends of CSER(t) and CPVR(t) as auxiliary constraints. Specifically, the time difference discrimination condition for the crack initiation stage is replaced by a theoretically calculated value based on rock mass physical parameters, replacing a fixed, manually assigned threshold. This ensures that the stage discrimination conditions are physically constrained rather than artificially defined. This difference makes the stage discrimination results more stable, effectively avoiding misjudgments caused by instantaneous fluctuations in single parameters, and eliminating dependence on laboratory test calibration parameters.

[0134] Key Difference 4: Stability Criterion for Continuous Sliding Windows

[0135] Existing technologies often result in stage discrimination errors due to transient noise. This invention addresses this by establishing a stability constraint that "a discrimination result can only be officially output when the criteria for n consecutive sliding windows all point to the same stage," filtering out random noise through continuous verification over time. This mechanism replaces the manual requirement of a fixed number of windows, giving the stability criterion a probabilistic statistical basis and further improving the reliability and engineering usability of the discrimination output.

Claims

1. A method for determining the timing of acoustic emission and fracturing pressure during the hydraulic fracturing stage of rock mass: characterized in that, Includes the following steps: S1. Multi-source data synchronous acquisition: Acoustic emission sensors and fracturing pressure sensors are simultaneously deployed in the hydraulic fracturing operation area of ​​deep resource mining to collect rock mass acoustic emission signals and fracturing pipeline pressure data, and to synchronize the two types of data in time. S2. Calculation and cumulative smoothing of acoustic emission energy release rate: Divide the acoustic emission signal into segments according to a fixed time window length Δt0, calculate the acoustic emission energy and acoustic emission energy release rate SER(i) in each time window, and use a fixed length sliding window to perform cumulative smoothing on SER(i) to construct the cumulative acoustic emission energy release rate curve CSER(t). S3. Calculation and cumulative smoothing of fracturing pressure change rate: Divide the fracturing pressure data into segments according to the same time window length Δt0 as in step two, calculate the fracturing pressure change rate PVR (i) in each time window, and use a sliding window with the same parameters as in step two to perform cumulative smoothing on PVR (i) to construct the cumulative fracturing pressure change rate curve CPVR (t). S4. Characteristic Inflection Point Identification of Two-Parameter Cumulative Curve: Identifying the peak inflection point time t of the CPVR(t) curve. P Simultaneously identify the sudden inflection point t of the CSER(t) curve. E ; S5. Energy-Pressure Response Time Difference Calculation: Based on the formula Calculate the response time difference between the inflection point of the sudden increase in acoustic emission energy and the inflection point of the peak fracturing pressure. ; S6. Hydraulic fracture stage identification: Combining the morphological evolution characteristics, inflection point order, and response time difference of the CPVR(t) and CSER(t) curves. The hydraulic fracturing process of rock mass is classified into corresponding stages in the compaction stage, fracturing initiation stage, propagation and penetration stage, and pressure stabilization and diffusion stage. S7. Result Verification, Output, and Process Feedback: Set continuous sliding window stability verification rules. After successful verification, the output includes the current stage, stage time, duration, CSER(t), CPVR(t), and response time difference. The judgment results are included; at the same time, the judgment results are fed back to the hydraulic fracturing process control system in real time to adjust the injection rate, pump pressure limit and / or pump shutdown timing.

2. The method for determining the acoustic emission-fracturing pressure timing during the hydraulic fracturing stage of rock mass according to claim 1, characterized in that, In step s2, the formula for calculating the acoustic emission energy within the i-th time window is: In the formula, N i A is the ringing count of acoustic emission events within the i-th time window; j Given the peak amplitude of the j-th acoustic emission event, calculate the acoustic emission energy release rate SER(i) for the i-th time window: Calculate the acoustic emission energy release rate SER(i) for the i-th time window: ， The formula for calculating the cumulative acoustic emission energy release rate corresponding to the k-th sliding window, given a sliding window length of W, is as follows: 。 3. The method for determining the acoustic emission-fracturing pressure timing during the hydraulic fracturing stage of rock mass according to claim 1, characterized in that, In step s3, the formula for calculating the rate of change of fracturing pressure within the i-th time window is: ， In the formula, Let fracturing pressure be the value at the end of the i-th time window; For the first The fracturing pressure value at the end of each time window, where Δt0 is the fixed time window length; The formula for calculating the cumulative fracturing pressure change rate corresponding to the kth sliding window is: 。 4. The method for determining the acoustic emission-fracturing pressure timing during the hydraulic fracturing stage of rock mass according to claim 1, characterized in that, The peak inflection point of the CPVR(t) curve in s4. satisfy: ， Right now The moment when the first derivative of the curve changes from positive to negative and the second derivative becomes negative indicates that the driving force of the fracturing pressure growth changes from dominant to attenuated, corresponding to the critical state of the rock mass that is about to or has already begun hydraulic fracturing. The sudden inflection point of the CSER(t) curve satisfy: ， The current slope of the CSER(t) curve is greater than the average slope of the previous baseline. The value is multiplied by a factor of m, and this state persists for at least m sliding windows, where: Let be the average slope of the baseline segment of the CSER(t) curve before time t; The threshold value is the sudden increase factor, and m is the preset duration window number.

5. The method for determining the acoustic emission-fracturing pressure timing during the hydraulic fracturing stage of rock mass according to claim 1, characterized in that, In step s6 The specific stage discrimination criteria are as follows: During the compaction phase: CPVR(t) continues to increase monotonically, CSER(t) is below the preset baseline threshold S1 and remains stable, and no valid t is identified. P With t E Inflection point, or Exceeds the valid discrimination range; Initiation stage: CPVR(t) passes through the peak inflection point t P Afterwards, it turns downward, and CSER(t) shows a sudden increase at the inflection point t. E And satisfy 0 < Δt < Δt theory ·(1+γ), where Δt theory γ is the theoretical time difference, and γ is the allowable deviation coefficient; During the extended and connected phase: CSER(t) remains in the high range above the threshold S2 or experiences a second sudden increase, CPVR(t) continues to decrease or fluctuates slightly, and Δt tends to stabilize; During the voltage stabilization and diffusion stage: CSER(t) changes from a high level to a decrease, and CPVR(t) converges to the stabilizing voltage P. steady Furthermore, if the fluctuation amplitude is less than the preset threshold, Δt will no longer be used as the main criterion.

6. The method for determining the acoustic emission-fracturing pressure timing during the hydraulic fracturing stage of rock mass according to claim 1, characterized in that, The stability verification rule is as follows: when the stage judgment results of 3-5 consecutive sliding windows are consistent, the verification is deemed to have passed and the judgment result is officially output.

7. The method for determining the acoustic emission-fracturing pressure timing during the hydraulic fracturing stage of rock mass according to claim 1, characterized in that, In step s2, the acoustic emission energy release rate is replaced by the acoustic emission ringing count rate or the acoustic emission event amplitude change rate to complete the construction of the cumulative curve and the identification of the inflection point.

8. The method for determining the acoustic emission-fracturing pressure timing during the hydraulic fracturing stage of rock mass according to claim 1, characterized in that, Based on step s3, the fracturing injection rate data are collected synchronously, the injection rate change rate is calculated, and the cumulative injection rate change rate curve is constructed. Combined with CSER (t) and CPVR (t), a three-parameter time-series correlation discrimination system is formed.

Images