Comprehensive monitoring and treatment method for failure zone of deep unloading hard rock roadway under blasting disturbance

Abstract

This invention provides a comprehensive monitoring and treatment method for the surrounding rock damage area in deep unloaded hard rock tunnels under blasting disturbance, belonging to the field of rock mass engineering monitoring technology. This invention extracts the reference amplitude sequence, reference phase sequence, disturbance amplitude sequence, and disturbance phase sequence by performing a Hilbert transform on the radar reflected wave sequences before and after blasting disturbance; constructs an amplitude difference term and a phase amplitude interference term and performs nonlinear coupling to obtain a time-shifted differential signal sequence; performs a Hilbert-Huang transform on it, multiplies the frequency domain weighting term by the square of the time-varying amplitude and integrates to obtain the cumulative marginal energy; calculates the cumulative damage index by combining a fractional damping term and an exponential decay term; delineates the damage area based on the spatial distribution of the cumulative damage index exceeding a preset threshold at each collection point, and outputs the surrounding rock damage result by comparing the maximum value of the index within the damage area with a preset level interval, thereby achieving accurate quantitative definition of the hidden damage boundary of deep surrounding rock and efficient early warning of engineering disasters.

Description

Technical Field

[0001] This invention relates to the field of rock mass engineering monitoring technology, specifically a comprehensive monitoring and treatment method for the surrounding rock damage area in deep unloaded hard rock tunnels under blasting disturbance. Background Technology

[0002] With the development of deep underground engineering, deep hard rock tunnels are highly susceptible to micro-crack initiation and structural damage under the combined effects of excavation unloading and frequent blasting disturbances, leading to a deteriorated area with degraded physical and mechanical properties. Comprehensive and accurate monitoring of this damaged area is a prerequisite for ensuring construction safety. Traditional detection methods such as borehole observation or acoustic testing have drawbacks, including limited sampling points, time-consuming and labor-intensive methods, and difficulty in large-scale simultaneous observation. In recent years, ground-penetrating radar (GPR) has become the mainstream method for surrounding rock monitoring due to its macroscopic advantages of non-destructive continuous detection. However, the internal environment of deep unloaded hard rock is extremely complex, and traditional single-radar observation lacks deep integration of multi-dimensional physical parameters of electromagnetic waves, making it difficult to objectively reflect the true damage evolution state of the surrounding rock when dealing with complex rock masses.

[0003] Existing ground-penetrating radar (GPR)-based rock monitoring methods typically collect radar reflection data from the target area and perform basic filtering to eliminate background noise and high-frequency interference from shallow environments. However, a fatal flaw of these existing technologies lies in their over-reliance on fixed thresholds set by expert experience for simple waveform comparisons during feature extraction. Furthermore, due to the limitations of single-parameter evaluation, they often necessitate extensive on-site rock core sampling tests to serve as a true benchmark for assessing the damage state. Overall, these GPR-based rock monitoring methods primarily rely on a combined approach of extracting single radar amplitude attenuation features and supplementing with physical sampling data to determine the damaged area.

[0004] In actual deep unloading environments, the microscopic fractures within the rock mass caused by blasting disturbance often result in intertwined distortions of multiple physical parameters, such as amplitude and phase, in the radar reflection wave sequence. However, existing radar signal processing methods struggle to decouple and balance the physical interference between amplitude differences and phase deflections, leading to ambiguity in the physical meaning of the assessment results. Furthermore, deep hard rock often contains static geological noise such as natural isolated joints or localized water seepage. Existing evaluation methods are prone to misinterpreting such localized isolated interferences as continuous structural damage caused by blasting, resulting in drastic jumps in monitoring results. This makes it difficult to objectively and stably reflect the true degree of damage to the surrounding rock and the spatial continuity of the damaged area under blasting disturbance solely based on radar wave data.

[0005] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0006] The purpose of this invention is to provide a comprehensive monitoring and treatment method for the surrounding rock damage area in deep unloaded hard rock roadways under blasting disturbance, so as to solve the problems mentioned in the background art.

[0007] To achieve the above objectives, the present invention provides the following technical solution: A comprehensive monitoring and treatment method for the surrounding rock damage area in deep unloaded hard rock roadways under blasting disturbance, the specific steps of which include: Step 1: For the acquisition point of the deep unloaded hard rock roadway, collect the radar reflection wave sequence of the surrounding rock before and after the blasting disturbance within the preset time window. Time-align the two sequences and perform Hilbert transform on them respectively. Reconstruct the corresponding reference analytical sequence and disturbance analytical sequence of the acquisition point. Extract the reference amplitude sequence and reference phase sequence of the acquisition point from the reference analytical sequence. Extract the disturbance amplitude sequence and disturbance phase sequence of the acquisition point from the disturbance analytical sequence. Step 2: For each time node of the preset time window, construct an amplitude difference term based on the energy variation characteristics of the perturbation amplitude sequence and the reference amplitude sequence. Combine the deflection characteristics of the perturbation phase sequence and the reference phase sequence and the corresponding double amplitude characteristics to construct a phase amplitude interference term and nonlinearly couple it with the amplitude difference term and arrange it in time sequence to obtain the time-shifted differential signal sequence of the acquisition point. Step 3: Perform Hilbert-Huang transform on the time-shifted differential signal sequence to decompose the intrinsic mode function components, extract the time-varying amplitude and time-varying frequency of the target component at each time node, construct the frequency domain weighting term of the corresponding node based on the ratio feature of the time-varying frequency and the preset radar center frequency, multiply it by the square of the time-varying amplitude, perform sequential summation within the preset time window, and obtain the cumulative marginal energy of the acquisition point. Step 4: Construct a fractional damping term based on the ratio of the square of the cumulative marginal energy to the preset reference energy, construct an exponential decay term based on the negative exponential decay characteristic of the ratio of the cumulative marginal energy to the preset reference energy, and multiply it with the fractional damping term to obtain the cumulative damage index of the sampling point. Step 5: Delineate the damaged area based on the spatial continuous distribution characteristics of the cumulative damage index exceeding the preset threshold at each collection point, and compare the maximum value of the cumulative damage index within the damaged area with the preset level interval to output the surrounding rock damage result.

[0008] Furthermore, the steps for defining the collection point and the preset time window specifically include: On the exposed surrounding rock surface of the deep unloading hard rock tunnel, longitudinal survey lines are planned along the axial direction of the tunnel, and transverse survey lines are planned along the cross-sectional contour of the tunnel. The detection grid line formed by the interweaving of the longitudinal survey lines and the transverse survey lines is defined as the line to be measured, and multiple collection points are set at equal intervals on the line to be measured. For any collection point on the line to be measured, determine the upper and lower boundaries of the preset monitoring depth of the target surrounding rock, and obtain the propagation speed of radar electromagnetic waves in the current surrounding rock mass; Based on the physical conversion relationship between depth and two-way propagation time, the upper and lower boundaries are converted into corresponding time start and time end points, and the interval between the time start and time end points is defined as a preset time window.

[0009] Furthermore, the steps of time-aligning the two sequences, reconstructing the analytical sequence, and extracting the amplitude and phase sequences specifically include: By extracting the starting point of the direct wave, time zero-point alignment is performed on the radar reflected wave sequences before and after the blasting disturbance; When performing the Hilbert transform, the radar reflection wave sequence before the blast disturbance, after time zero-point alignment, is mapped from the one-dimensional real number domain to the reference analytical sequence of the acquisition point on the complex plane, and the radar reflection wave sequence after the blast disturbance, after time zero-point alignment, is mapped from the one-dimensional real number domain to the disturbance analytical sequence of the acquisition point on the complex plane. Calculate the modulus sequence and complex argument sequence of the reference analytical sequence in the complex plane, and extract them as the reference amplitude sequence and reference phase sequence of the acquisition point, respectively. Calculate the modulus sequence and complex argument sequence of the perturbation analytical sequence in the complex plane, and extract them as the perturbation amplitude sequence and perturbation phase sequence of the acquisition point, respectively.

[0010] Furthermore, the step of constructing the amplitude difference term based on the energy variation characteristics of the perturbation amplitude sequence and the reference amplitude sequence specifically includes: For any time point within the preset time window: Calculate the numerical difference between the perturbation amplitude sequence and the reference amplitude sequence at this time node and label it as the first difference. Calculate the square of the first difference and construct an amplitude difference term that characterizes the absolute change of reflected energy at this time node. Calculate the phase difference between the perturbation phase sequence and the reference phase sequence at this time node and label it as the second difference value. Calculate half of the second difference value and take its corresponding square sinusoidal value to construct the deflection feature at this time node. Calculate the product of the perturbation amplitude sequence and the reference amplitude sequence at this time node, and take four times the value of the product to construct the dual amplitude feature at this time node; By multiplying the dual-amplitude feature and the deflection feature at this time point, the phase-amplitude interference term characterizing waveform delay and distortion amplification is obtained. The amplitude difference term and phase interference term at this time node are summed, and the square root operation is performed on the summation result to achieve nonlinear coupling. The time-shift differential signal value corresponding to this time node is calculated. All time nodes within the preset time window are traversed, and the time-shift differential signal values ​​at all time nodes are extracted and combined in chronological order to form the time-shift differential signal sequence of this acquisition point. The specific calculation formula is as follows: In the formula, For the first The time-shifted differential signal values ​​calculated at each time point and The perturbation amplitude sequence and the reference amplitude sequence are respectively at the 1st... Values ​​at each time point and The perturbation phase sequence and the reference phase sequence are respectively at the 1st... Values ​​at each time point This refers to the time node number.

[0011] Furthermore, the specific steps for obtaining the cumulative marginal energy at the collection point are as follows: The time-shifted differential signal sequence is characterized as a non-stationary signal bearing the characteristics of surrounding rock damage. The Hilbert-Huang transform is used to adaptively decompose it and map it to the instantaneous time-frequency domain to deconstruct the dynamic frequency mutation and energy fluctuation induced by rock mass fracture, thereby extracting the time-varying amplitude and time-varying frequency at each time node. For any given time point, calculate the ratio of the time-varying frequency to the preset radar center frequency at that time point. Summate this ratio with the value 1 and extract the natural logarithm of the sum to construct the frequency domain weighting term representing the nonlinear gain of high-frequency micro-damage corresponding to that time point. The instantaneous energy corresponding to the time node is obtained by multiplying the frequency domain weighting term corresponding to the time node with the square of the time-varying amplitude at the time node. The instantaneous energy value corresponding to each time node is multiplied by a preset sampling time interval, and the product results corresponding to all time nodes within the preset time window are sequentially summed to calculate the cumulative marginal energy of the sampling point. The specific calculation formula is as follows: In the formula, To calculate the cumulative marginal energy at this collection point, The total number of sampling points within the preset time window. The time node number is within the range of , For the first The time-varying amplitude corresponding to each time point To target time-shifted differential signals within a preset time window, the single dominant intrinsic mode function component with the highest total energy proportion within the entire time window, selected through empirical mode decomposition, is located at a specific time node. The time-varying frequency below, To preset the radar center frequency, This is the preset sampling time interval.

[0012] Furthermore, the specific steps for obtaining the cumulative damage index at the sampling point are as follows: Extract the cumulative marginal energy at the collection point and calculate the square of the cumulative marginal energy; Calculate the square of the preset reference energy, and sum the square of the cumulative marginal energy with the square of the preset reference energy to obtain a first sum value; Calculate the ratio of the square of the accumulated marginal energy to the first sum, and construct a fractional damping term corresponding to the acquisition point to characterize the numerical suppression effect in the early stage of damage; Calculate the ratio of the accumulated marginal energy to the preset reference energy and calibrate it as the first ratio. Extract the decay value with the natural constant as the base and the opposite of the first ratio as the exponent. Calculate the difference between the value 1 and the decay value to construct the exponential decay term corresponding to the acquisition point that characterizes the late-stage convergence boundary of the damage. The cumulative damage index of the acquisition point is calculated by multiplying the fractional damping term and the exponential decay term corresponding to that acquisition point. The specific calculation formula is as follows: In the formula, To calculate the cumulative damage index for this sampling point, This represents the cumulative marginal energy at the collection point. This is the preset reference energy for undisturbed, intact rock masses.

[0013] Furthermore, the specific planning steps for the damaged area include: The spatial coordinates of each sampling point on the line to be tested are obtained, and the cumulative damage index corresponding to each sampling point is compared with the preset threshold. Abnormal collection points that have a cumulative damage index exceeding the preset threshold are selected. Based on the spatial coordinates, consecutively adjacent abnormal collection points in space are clustered into the same set, and the physical spatial boundary corresponding to the set is defined as an independent damage area.

[0014] Furthermore, the specific steps for outputting the surrounding rock damage results are as follows: For any damaged area, extract the maximum value of the cumulative damage index of each abnormal collection point in the set corresponding to that damaged area; The maximum value is compared with a preset level range, which consists of a progressively increasing range of mild damage, moderate damage, and severe damage. When the maximum value falls within the minor damage range, it is determined that the damaged area has suffered minor damage in this blasting disturbance. When the maximum value falls within the moderate damage range, it is determined that the damaged area has suffered moderate damage in this blasting disturbance. When the maximum value falls within the severe damage range, the damaged area is determined to have suffered severe damage in this blasting disturbance.

[0015] Compared with the prior art, the beneficial effects of the present invention are: This invention extracts amplitude and phase sequences from radar reflected wave sequences using Hilbert transform to construct phase-amplitude interferometry terms, and nonlinearly couples these terms with amplitude difference terms. Subsequently, time-varying frequencies are extracted using Hilbert-Huang transform to construct frequency-domain weighting terms. Combined with time-domain integrals and damping attenuation functions, the cumulative damage index is calculated. This invention overcomes the shortcomings of traditional single radar amplitude observation, which is prone to missed detections and highly dependent on physical core sampling. By deeply physical fusing energy attenuation characteristics and waveform distortion characteristics, it accurately captures and amplifies the electromagnetic wave variation response induced by micro-fractures within the rock mass, thereby significantly improving the sensitivity of damage monitoring. This invention also utilizes the spatially continuous distribution characteristics of the cumulative damage index exceeding a preset threshold at each acquisition point to cluster and delineate spatially adjacent abnormal acquisition points into independent damage areas, ultimately outputting the surrounding rock damage results. This mechanism effectively eliminates static geological noise interference caused by natural isolated joints or localized random seepage in deep hard rock, preventing the misjudgment of localized isolated numerical anomalies as continuous structural damage caused by blasting. Ultimately, driven purely by radar data, it achieves highly interference-resistant, stable, and accurate determination of the true degree of damage and spatial failure boundaries of the surrounding rock in deep unloading tunnels. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of the overall method flow of the present invention; Figure 2 A schematic diagram of the nonlinear mapping combination of cumulative marginal energy and cumulative damage index at the acquisition point; Figure 3 This is a feature map of the first and second derivatives of the cumulative damage index. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0018] It should be noted that, unless otherwise defined, the technical or scientific terms used in this invention should have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in this invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0019] Example: Please see Figures 1-3 The present invention provides a technical solution: A comprehensive monitoring and treatment method for the surrounding rock damage area in deep unloaded hard rock roadways under blasting disturbance, the specific steps of which include: Step 1: For the acquisition point of the deep unloaded hard rock tunnel, collect the radar reflection wave sequence of the surrounding rock before and after the blasting disturbance within the preset time window. Time-align the two sequences and perform Hilbert transform on them respectively. Reconstruct the corresponding reference analytical sequence and disturbance analytical sequence of the acquisition point. Extract the reference amplitude sequence and reference phase sequence of the acquisition point from the reference analytical sequence. Extract the disturbance amplitude sequence and disturbance phase sequence of the acquisition point from the disturbance analytical sequence.

[0020] Because the surrounding rock is under high stress unloading after tunnel excavation, the damage caused by blasting disturbance exhibits significant spatial non-uniformity within the rock mass. Single-point monitoring data is highly susceptible to random interference from local rock property fluctuations, leading to judgment bias. Therefore, in this embodiment, longitudinal survey lines are planned along the axial direction of the exposed surrounding rock surface of the deep unloaded hard rock tunnel, and transverse survey lines are planned along the cross-sectional contour of the tunnel. The longitudinal and transverse survey lines are interwoven to form a detection grid covering the area to be monitored. This detection grid line is collectively defined as the line to be measured, thereby establishing a comprehensive perception boundary for the evolution of damage to the tunnel roof and sidewalls in a spatial dimension.

[0021] In this embodiment, to achieve a smooth transformation from discrete sampling data to continuous spatial damage characteristics, multiple sampling points are equidistantly set on the test line, with a spacing of 0.5 meters between the sampling points and a total of 50 sampling points. This spacing and number setting aims to ensure the analytical accuracy of the detection grid in complex rock mass environments, ensuring that the spatial distribution patterns of minute fractures within the surrounding rock can be captured subsequently. For each sampling point, the target rock mass depth needs to be precisely determined through a preset time window. This embodiment sets the preset monitoring depth range for the target surrounding rock to 0.5 meters to 3.0 meters. The aim is to optimize the detection signal-to-noise ratio and target focusing through physical depth range constraints. Setting the initial depth to 0.5 meters effectively shields shallow air-coupled interference waves and high-energy direct waves, preventing their strong signal characteristics from masking near-surface micro-fracture echoes. Locking the upper limit at 3.0 meters accurately filters irregular clutter and background noise generated by deep, irrelevant rock masses, allowing subsequent analytical algorithms to completely eliminate interference from invalid echoes. This ensures that monitoring data is accurately focused on the core unloading area of ​​the blasted loosening zone, significantly improving the reliability of surrounding rock damage feature extraction and the analytical accuracy of boundary delineation. Utilizing the propagation speed of radar reflected waves in the current hard rock mass, where the relative permittivity of deep hard rock is typically between 6 and 9, the propagation speed is set to... This ensures absolute accuracy when converting spatial depth to the time axis. Based on the physical conversion relationship between depth and two-way propagation time, the corresponding time starting point (approximately) is calculated from the depth range. ) and the end of the time (approximately The time interval defined in this way can perfectly cover the preset monitoring depth of the target surrounding rock from 0.5 meters to 3.0 meters, completely filtering out direct waves and background noise from extremely deep layers. The preset time window is defined by this interval. The physical significance of setting this time window is to eliminate interference from shallow air-coupled waves and clutter signals from deep irrelevant rock masses, so that subsequent analysis can accurately focus on the core unloading area that may exist in the blasting loosening zone.

[0022] This embodiment uses a preset radar center frequency of... The ground-penetrating radar antenna sets the sampling frequency to And set the preset sampling time interval to The high center frequency combined with an extremely high sampling frequency ensures that the acquired signal has sufficient time resolution, thereby capturing the microsecond-level high-frequency transient changes caused by blasting disturbances. For each acquisition point, two key acquisitions are performed within a preset time window: one for the baseline state of the surrounding rock before blasting disturbance, and the other for the disturbance state after blasting disturbance. To eliminate time axis jitter caused by slight differences in antenna placement, trigger delay, or electronic drift during the two detection processes, this embodiment extracts the direct wave start point as a reference and performs time zero-point alignment on the radar reflection wave sequences before and after blasting. This processing ensures an absolute one-to-one correspondence between the data nodes of the two sequences on the time axis.

[0023] At the signal analysis level, a Hilbert transform is performed on the radar reflection wave sequence after time zero-point alignment. The mathematical essence of this Hilbert transform is to map the wave signal, originally located in the one-dimensional real domain and only capable of expressing oscillation amplitude, to the complex plane. By superimposing the real and imaginary orthogonal signals, a reference analytical sequence and a perturbation analytical sequence containing complex characteristics for the acquisition point are reconstructed. This reconstruction reveals the instantaneous characteristics of the signal at any microscopic moment. Subsequently, by calculating the modulus sequence of the analytical sequence in the complex plane, the reference amplitude sequence and the perturbation amplitude sequence for the acquisition point are extracted to characterize the intensity evolution of the reflected energy envelope. Simultaneously, the complex argument sequence of the analytical sequence in the complex plane is calculated to extract the reference phase sequence and the perturbation phase sequence for the acquisition point. Phase extraction aims to capture the microscopic deflection and propagation delay characteristics of the waveform caused by changes in the rock mass damage medium. Through this multidimensional decoupling process, the original chaotic radar echo is transformed into analytical components with clear physical indication meaning.

[0024] Step 2: For each time node of the preset time window, construct an amplitude difference term based on the energy variation characteristics of the perturbation amplitude sequence and the reference amplitude sequence. Combine the deflection characteristics of the perturbation phase sequence and the reference phase sequence and the corresponding double amplitude characteristics to construct a phase amplitude interference term and nonlinearly couple it with the amplitude difference term and arrange it in time sequence to obtain the time-shifted differential signal sequence of the acquisition point.

[0025] In this embodiment, by performing numerical decoupling and nonlinear operations on each time node within a preset time window, the aim is to transform the weak wave field distortion caused by rock mass damage into a time-shifted differential signal with strong characteristic expressive power. For any time node within the preset time window, the absolute energy change characteristics and waveform deflection interference characteristics at that node are calculated respectively. This node-level traversal calculation logic can accurately map the evolution characteristics of radar reflected waves on the time axis to the physical spatial coordinates of the radial depth of the surrounding rock.

[0026] When constructing the amplitude difference term characterizing the absolute change in reflected energy, the numerical difference between the perturbation amplitude sequence and the reference amplitude sequence at that time node is calculated and squared to obtain the amplitude difference term characterizing the absolute change in reflected energy. The specific calculation formula is as follows: In the formula, For the first The values ​​corresponding to the amplitude difference term at each time point This refers to the time node number. and The perturbation amplitude sequence and the reference amplitude sequence are respectively at the 1st... The values ​​at each time point.

[0027] The amplitude difference term characterizes the degree of pure energy loss or surge of radar reflected waves at the same depth or time point after deep hard rock is disturbed by blasting. Its value increases quadratically with the increase of the absolute difference between the disturbance amplitude and the reference amplitude. This amplitude difference term uses squaring operation instead of simple absolute value of difference. Since the energy of the wave is proportional to the square of its amplitude, squaring operation not only eliminates the interference of alternating positive and negative signs of the wave amplitude, but also directly and objectively quantifies the absolute change of reflected energy caused by abrupt changes in the physical properties of the medium. In response to the large amount of background white noise mixed in the original radar signal, squaring operation cleverly constructs a nonlinear amplification and suppression mechanism, so that significant damage distortion signals (i.e., large differences) are amplified many times, while small random noise fluctuations (i.e., very small differences) are rapidly attenuated and suppressed, which greatly improves the signal-to-noise ratio of the extracted bottom layer damage features. Furthermore, a large amplitude difference term indicates that blasting disturbance causes numerous new micro-fractures in the surrounding rock of the tunnel or displacement of existing structural surfaces. The intrusion of air or groundwater alters the local dielectric constant of the rock mass, triggering strong scattering, reflection, or absorption of radar reflected waves. Conversely, a small amplitude difference term indicates that the blasting stress wave has not caused substantial physical damage to the rock mass in the area, and the density and continuity of the rock mass remain largely unchanged. Thus, the introduction of this amplitude difference term achieves, technically, the pure stripping and highly sensitive capture of actual physical damage energy.

[0028] When constructing the phase-amplitude interferometry term, the waveform delay and distortion amplification effects must be considered comprehensively. First, the phase difference between the perturbation phase sequence and the reference phase sequence at the given time node is calculated, and the square of half this phase difference is taken to construct the deflection characteristic at that time node. Then, the product of the perturbation amplitude sequence and the reference amplitude sequence at that time node is calculated and its value is quadrupled to construct the double-amplitude characteristic at that time node. Multiplying this double-amplitude characteristic with the deflection characteristic yields the phase-amplitude interferometry term characterizing the waveform delay and distortion amplification. The specific calculation formula is as follows: In the formula, For the first The values ​​corresponding to the phase amplitude interference terms at each time point This refers to the time node number. and The perturbation phase sequence and the reference phase sequence are respectively at the 1st... The values ​​at each time point.

[0029] The phase amplitude interference term characterizes the changes in the propagation path of radar reflected waves caused by the misalignment of the internal structure and the opening of fissures in deep hard rock after blasting disturbance, as well as the waveform phase deflection and interference distortion effects caused by this. Its value is linearly positively correlated with the product of the amplitudes before and after the disturbance, and nonlinearly positively correlated with the increase of the phase difference between the disturbance phase and the reference phase. This phase-amplitude interferometry term uses the product of amplitudes before and after the disturbance as a weighting coefficient to modulate the half-angle sine square of the phase difference, rather than solely considering the absolute phase shift. In the complex electromagnetic environment of deep unloaded hard rock, simple high-frequency background noise or minute seepage can easily cause phase jumps without physical meaning. The amplitude-product modulation mechanism not only follows the classical interference principle of complex vector waves, but also constructs a rigid constraint barrier of amplitude-controlled phase from the algorithm's bottom layer. To address the shortcomings of traditional single phase difference calculation, which is prone to false abrupt changes and misjudgments, this nonlinear coupling formula ensures that the phase shift that occurs will only be judged and extracted as a real structural damage feature under the background of radar reflected waves with sufficient energy reflection, thereby filtering out phase artifacts in weak energy backgrounds. Furthermore, when the phase-amplitude interference term is large, it means that the surrounding rock has not only undergone microscopic fracturing, but also macroscopic structural displacements such as rock layer slippage, faulting, or unloading expansion. Changes in the thickness or distance of the physical medium cause substantial shifts in the electromagnetic wave flight time, resulting in significant phase interference. When the phase-amplitude interference term is small, it means that the rock mass structure is relatively stable, with no obvious physical displacement or large-scale crack opening, and the radar reflected wave propagation path remains highly consistent. By utilizing the wavefield interference principle, the microsecond-level waveform delay and amplitude distortion caused by rock mass structural anomalies were compositely characterized, enabling the keen detection of very early signs of micro-damage that have not yet caused large changes in macroscopic amplitude but have already led to wavefront phase lag.

[0030] Finally, the amplitude difference term and phase interference term at this time point are summed, and the square root operation is performed on the summation result to achieve nonlinear coupling, thus calculating the time-shifted differential signal value corresponding to this time point. This allows for deep fusion of the variation in the energy intensity dimension and the shift in the waveform phase dimension. By utilizing the wavefield interference principle, the absolute loss of energy intensity and the hysteresis deflection of the wavefront phase are deeply fused, thereby significantly improving the sensitivity to identify very early and weak damage characteristics of the surrounding rock in complex background interference. The specific calculation formula is as follows: In the formula, For the first The time-shifted differential signal values ​​calculated at each time point For amplitude difference, and The perturbation amplitude sequence and the reference amplitude sequence are respectively at the 1st... Values ​​at each time point For phase-amplitude interference terms, This represents the deflection characteristics at this point in time. This represents the double-amplitude characteristic at this time point. and The perturbation phase sequence and the reference phase sequence are respectively at the 1st... Values ​​at each time point This refers to the time node number.

[0031] Among them, the time-shift differential signal value corresponding to each time node represents the degree of absolute comprehensive distortion of the physical state of the rock mass at a specific depth or time node after the surrounding rock of the deep hard rock tunnel has undergone blasting unloading. Its value is jointly determined by the amplitude difference term and the phase amplitude interference term and is positively correlated. The calculation of this time-shift differential signal sequence is not a simple linear splicing or scalar addition of amplitude and phase characteristics, but rather an upscaling mapping of the radar signal from the one-dimensional time domain to a two-dimensional complex vector phasor space. Nonlinear physical coupling is achieved by calculating the Eulerian distance between the radar reflected wave vectors before and after the disturbance. Since the blasting damage in the deep unloaded rock mass has multi-dimensional characteristics of energy dissipation from micro-fractures and macroscopic displacement of structural surfaces, the underlying logic developed through the complex vector cosine theorem not only achieves perfect decoupling and unification of the two orthogonal physical dimensions of energy mutation characteristics and travel time offset characteristics, and constructs a dual anti-interference verification barrier with amplitude-controlled phase and phase-positive amplitude, but also its externally nested square root operation performs forced dimensionality reduction and normalization processing on the dimensions of the internal energy level, making it maintain a linear mapping relationship with the original radar electric field amplitude of the underlying layer with the same dimensions, thus providing a dimensionless standard physical base for the subsequent smooth extraction of accumulated marginal energy. Furthermore, when the value of the time-shift differential signal sequence is extremely large (i.e., a significant distortion peak appears), it means that the surrounding rock has not only generated a large number of fissures at the microscopic level, causing electromagnetic wave energy to be strongly absorbed or scattered, but also that substantial structural slippage or tensile cracking has occurred at the macroscopic level, causing abrupt changes in the propagation path. This directly reflects that the rock mass in this area has approached or entered the plastic failure zone or even the complete collapse zone. When the value of the time-shift differential signal sequence is extremely small or close to zero, it means that the blasting stress wave passes through the area only in the form of elastic waves, and the reflection energy and propagation path of the radar reflected wave have not undergone substantial changes. The monitored area still belongs to the structurally dense, non-damaged elastic zone. Therefore, the nonlinear extraction mechanism of the time-shift differential signal value completely removes non-damaging background clutter and system instrument drift in deep and complex geological environments, improving the accuracy of the final surrounding rock stability early warning criterion.

[0032] Step 3: Perform Hilbert-Huang transform on the time-shifted differential signal sequence to decompose the intrinsic mode function components, extract the time-varying amplitude and time-varying frequency of the target component at each time node, construct the frequency domain weighting term of the corresponding node based on the ratio feature of the time-varying frequency and the preset radar center frequency, multiply it by the square of the time-varying amplitude, perform sequential summation within the preset time window, and obtain the cumulative marginal energy of the acquisition point.

[0033] In this embodiment, due to the extremely complex internal fractures of hard rock caused by both blasting disturbance and deep high-stress unloading, the radar echo is a typical highly non-stationary and nonlinear signal. Traditional Fourier transform can only provide a global average spectrum and cannot characterize the frequency abrupt changes occurring at a specific depth (i.e., a specific time). Therefore, this embodiment performs Hilbert-Huang transform on the time-shifted differential signal sequence. First, empirical mode decomposition is performed on the time-shifted differential signal sequence to adaptively obtain multiple intrinsic mode function components at different time scales. This process aims to deconstruct the signal step-by-step from macroscopic structural response to microscopic fracture details based on the local characteristics of the signal. To eliminate random high-frequency electromagnetic interference generated by the startup of mining equipment in deep mines, this embodiment uses an energy correlation evaluation criterion to select the dominant intrinsic mode function component (IMF) with the highest correlation to rock mass damage evolution characteristics and the largest energy proportion from multiple intrinsic mode function components obtained through decomposition. Subsequently, only this dominant IMF component is subjected to Hilbert transform to accurately extract the instantaneous envelope amplitude and instantaneous frequency at each time point. Extracting the time-varying amplitude is to capture the instantaneous energy envelope intensity of electromagnetic waves generated by abrupt changes in the physical properties of the medium at the corresponding rock mass depth, while extracting the time-varying frequency is to sensitively perceive the high-frequency scattering and modulation effects of rock microfracture development on radar reflected waves.

[0034] In complex hard rock media, tiny newly formed fracture surfaces often cause strong scattering of electromagnetic waves in specific bands, resulting in a significant deviation of the instantaneous frequency of the echo from the source frequency. For any given time point, this embodiment calculates the ratio of the time-varying frequency at that time point to a preset radar center frequency. This ratio is then summed with a numerical value of 1, and its natural logarithm is extracted to construct a frequency domain weighting term representing the nonlinear gain of high-frequency micro-damage at that time point. The specific calculation formula is as follows: In the formula, For the first The values ​​corresponding to the frequency domain weighting terms at each time point. This refers to the time node number. To target time-shifted differential signals within a preset time window, the single dominant intrinsic mode function component with the highest total energy proportion within the entire time window, selected through empirical mode decomposition, is located at a specific time node. The time-varying frequency below, To preset the radar center frequency, in this embodiment, the preset radar center frequency is set to... .

[0035] The frequency domain weighting term characterizes the intensity of the high-frequency scattering and modulation effects of micro-fractures within deep rock masses on radar reflected waves. Its value is positively correlated with the rate of shift of the instantaneous frequency relative to the reference center frequency. Furthermore, due to the external nesting of a natural logarithmic function, this positive drive exhibits a nonlinear characteristic of initial sensitivity followed by a smoothing out of the later stages. This frequency domain weighting term uses the relative frequency shift rate and a nested natural logarithmic function for modulation, rather than employing a simple linear proportional multiplication. Since the newly formed microfractures generated in the early stage of deep blasting unloading are often much smaller than the radar wavelength, they are prone to strong Rayleigh scattering, which causes the echo signal to shift significantly to higher frequencies. Introducing a relative frequency offset rate can convert the pure microscopic physical frequency mutation into a gain weight of energy integral. Moreover, the complex environment of deep mines is prone to the intrusion of transient high-frequency spike pulses caused by the start-up of large mining equipment. If linear weighting is used, it will inevitably lead to an exponential explosion of accumulated energy. The natural logarithmic function is extremely sensitive to real, low-to-medium degree microfracture frequency shifts, but it will forcibly flatten its gain curve for extreme high-frequency electromagnetic distortion interference. In addition, the addition operation ensures that when there is no substantial frequency shift at the node, the weight term naturally and smoothly returns to zero, completely avoiding the singularity collapse of logarithmic calculation at the mathematical level. Furthermore, when the value of this frequency domain weighting term is large, it means that the radar wave encountered dense newly formed fracture surfaces when traversing the rock mass at this depth. The physical scale of the micro-fractures triggered strong high-frequency scattering, reflecting that the micro-damage evolution within the surrounding rock in this area was extremely active. When the value of this frequency domain weighting term is small or close to zero, it means that the frequency of the radar reflected wave remained basically near the fundamental frequency, without significant high-frequency abrupt changes. By quantifying the high-frequency distortion of the waveform using the frequency ratio and leveraging the nonlinear evolution characteristics of the logarithmic function, a significant amplification gain is given to the high-frequency abrupt changes characterizing early micro-damage. At the same time, its later convergence effectively suppresses the unlimited expansion of external extreme high-frequency noise, ensuring the objectivity and stability of the characteristic amplification.

[0036] The instantaneous energy at a given time point is obtained by multiplying the frequency domain weighted term corresponding to that time point with the square of the time-varying amplitude at that time point, thus achieving a cross-fusion of the base energy and the degree of frequency domain variation. The total number of sampling points within the preset time window and the preset sampling time interval of the radar equipment are obtained. The instantaneous energy value corresponding to each time point is multiplied with the preset sampling time interval, and the product results corresponding to all time points within the preset time window are sequentially summed to calculate the cumulative marginal energy of the sampling point. The cumulative marginal energy physically discretizes and accumulates the fleeting complex frequency domain fluctuations at microscopic time points along the propagation depth of the radar reflected wave, thereby reducing its dimensionality and transforming it into a stable and intuitive macroscopic scalar. This scalar is precisely equivalent to the sum of radar reflected wave energy scattering and variation induced by damage during blasting in the rock mass region where the sampling point is located. The specific calculation formula is as follows: In the formula, To calculate the cumulative marginal energy at this collection point, The total number of sampling points within the preset time window. The time node number is within the range of , For the first The time-varying amplitude corresponding to each time point To target time-shifted differential signals within a preset time window, the single dominant intrinsic mode function component with the highest total energy proportion within the entire time window, selected through empirical mode decomposition, is located at a specific time node. The time-varying frequency below, For frequency domain weighting terms, To preset the radar center frequency, in this embodiment, the preset radar center frequency is set to... , In this embodiment, the preset sampling time interval is set to... .

[0037] The cumulative marginal energy characterizes the absolute cumulative amount of effective damage scattering energy of radar electromagnetic waves caused by the development of micro-fractures and macro-structural damage within the surrounding rock mass within a set time window. Its value is determined by the square term of the instantaneous time-varying amplitude, the frequency domain weighting term, and the time integral increment, and they are positively correlated. This cumulative marginal energy does not adopt the conventional method of integrating the square of the amplitude alone in traditional radar technology, but instead deeply embeds the frequency domain weighting term within the discrete energy integral model. Since natural groundwater enrichment zones or smooth bedding in mines can also induce strong radar wave reflections, leading to amplitude distortion, direct integration would inevitably result in severe false damage misjudgments. Therefore, only when the radar reflected wave simultaneously encounters a physical interface that produces strong reflection (i.e., large amplitude) and the interface is composed of rough, newly formed micro-cracks, thus inducing strong high-frequency scattering (i.e., large frequency deviation) will the reflected energy at the corresponding time period be recognized as the real marginal damage energy and accumulated, thereby completely eliminating artifact interference from non-damaging structures. The introduction of the logarithmic function in the frequency domain weighting term makes the contribution of extreme high-frequency interference to the total energy exhibit a diminishing marginal effect. Even if transient extreme high-frequency electromagnetic shocks generated by the start-up of electromechanical equipment are mixed into the monitoring environment, this integration system can forcibly suppress the exponential explosion of energy. In addition, through discrete summation operations with time intervals, the one-dimensional nanosecond-level electromagnetic wave travel time sequence collected by radar hardware was successfully and robustly reduced in dimension to be mapped into the spatial damage thickness intuitively required for geotechnical engineering disaster assessment. Therefore, when the cumulative marginal energy is extremely high, it means that throughout the entire process of the radar wave penetrating this depth range, it continuously encountered dense fracture surfaces and open fissures, directly reflecting that the rock mass support in the monitored area has undergone substantial degradation or even complete loss. When the cumulative marginal energy is extremely low or close to zero, it means that although the radar wave has normal propagation attenuation throughout the entire integration time window, it has not excited reflection energy with high-frequency deviation characteristics, indicating that the rock mass in this area is dense and the medium is continuous, and the blasting disturbance has not caused cumulative physical damage with engineering destructive significance. Thus, extracting the cumulative marginal energy provides a high-confidence physical quantitative benchmark that completely filters out geological environment and electromechanical transient artifacts for the subsequent construction of the final damage criterion and accurate delineation of the surrounding rock loosening zone failure boundary.

[0038] Step 4: Construct a fractional damping term based on the ratio of the square of the cumulative marginal energy to the preset reference energy, construct an exponential decay term based on the negative exponential decay characteristic of the ratio of the cumulative marginal energy to the preset reference energy, and multiply it with the fractional damping term to obtain the cumulative damage index of the sampling point.

[0039] In the early stages of rock fracturing, weak energy disturbances or stress redistribution are often absorbed by the structural strength of the rock mass itself, and do not necessarily lead to substantial rock mass deterioration. To quantify this physical process, this embodiment extracts the cumulative marginal energy at the sampling point and calculates its square. Simultaneously, it calculates the square of a preset reference energy, and then compares the square of the cumulative marginal energy with the sum of the squares of the two to construct the fractional damping term corresponding to that sampling point. The specific calculation formula is as follows: In the formula, For the fractional damping term, This represents the cumulative marginal energy at the collection point. In this embodiment, the preset reference energy is set to the value of the undisturbed, intact rock mass. .

[0040] The fractional damping term characterizes the activation threshold effect and saturation trend control of deep rock masses during the damage evolution process caused by blasting disturbance. Its value is generated by a nonlinear mapping of the relative magnitudes of the measured cumulative marginal energy and the preset reference energy. Instead of using a conventional linear ratio, this fractional damping term constructs a nonlinear fractional damping model based on the energy square term. At the level of rock mechanics and physics mapping, real deep hard rock inevitably possesses an elastic strain threshold dominated by the in-situ stress before macroscopic fracturing. This fractional damping term, using the energy square term, ingeniously constructs a nonlinear damping curve at the mathematical level, simulating the transition of rock mass from elastic deformation to plastic failure. In this embodiment, to shield against interference from the original rock background electromagnetic noise and elastic echo energy, a well-preserved rock mass far from the blasting zone is selected as the reference area at the far end of the roadway, and at least 10 repeated detections are performed to obtain multiple sets of reference echo sequences. The cumulative marginal energy of each sequence within a preset time window is calculated. After removing abnormal extreme values ​​caused by random electromagnetic pulses or local metallic foreign objects, the maximum value among the remaining effective cumulative marginal energy is extracted and multiplied by a reliability tolerance coefficient greater than 1. The result is used as the preset reference energy. The reliability tolerance coefficient, considering the conventional electromagnetic background characteristics of deep unloaded hard rock roadways and the thermal noise drift rate of radar equipment, has a range of values ​​within which... In this embodiment, the value is set to 1.15. In this embodiment, the preset reference energy value is set to... The preset reference energy serves as a physical benchmark for defining the transition of rock mass from an elastically stable state to a nonlinear evolution state with microfractures. It ensures effective suppression of background clutter and minor, damage-free fluctuations. In the initial stages of blasting disturbance, the strong denominator damping effect forcefully suppresses the increase in value, objectively reflecting the rock's elastic self-absorption characteristics of small energy fluctuations. This fractional damping term effectively filters out environmental disturbance energy that has not reached the critical point of damage, solving the false warning problem of traditional monitoring systems that judge damage based solely on energy fluctuations. Furthermore, when the value of this fractional damping term is small or close to 0, it means that the measured cumulative marginal energy has not yet reached the critical threshold for triggering large-scale damage. At this point, the incremental blasting energy is completely filtered or absorbed by the damping mechanism within this term, indicating that the rock mass still maintains its macroscopic structural integrity. When the value of this fractional damping term is large (i.e., it breaks through the damping inflection point and accelerates towards 1), it means that the measured cumulative marginal energy has overwhelmingly surpassed the preset reference energy representing the system benchmark, and the supporting framework inside the rock mass has been breached and completely crossed the nonlinear elastic threshold stage. The fractional damping term utilizes the mathematical properties of second-order homogeneous ratios to exhibit a strong numerical hysteresis effect in the initial stage of energy accumulation. This hysteresis effect effectively suppresses numerical hypersensitivity caused by local minor stress adjustments, and truly reflects the mechanical damping effect of intact rock mass resisting initial failure at the algorithm's underlying level.

[0041] In this embodiment, because the performance degradation of the internal structure of deep unloaded hard rock exhibits significant nonlinear characteristics during the disturbance and failure process, a simple linear energy mapping is insufficient to accurately reflect the deterioration state of the rock mass. Therefore, this embodiment calculates and calibrates the ratio of the accumulated marginal energy to the preset reference energy as the first ratio, extracts the attenuation value with a base of the natural constant and an exponent of the negative of the first ratio, and constructs the exponential attenuation term corresponding to the sampling point by calculating the difference between the numerical value 1 and the attenuation value. The specific calculation formula is as follows: In the formula, For the exponential decay term, This represents the cumulative marginal energy at the collection point. In this embodiment, the preset reference energy is set to the value of the undisturbed, intact rock mass. .

[0042] The exponential decay term characterizes the rate sensitivity of microscopic damage in deep rock mass to the accumulation of blasting energy, and its irreversible, gradual evolution towards the macroscopic damage limit. Its value is nonlinearly modulated by a negative exponential function of the ratio of the measured cumulative marginal energy to a preset reference energy. Instead of using a traditional linear accumulation model, this exponential decay term ingeniously constructs a negative exponential convergence model based on the energy ratio. At the level of rock mechanics and thermodynamics, rock fracture is essentially an irreversible energy dissipation process. Furthermore, according to the fundamental theory of continuous medium damage mechanics, the degree of rock mass damage inevitably has an absolute physical upper limit (i.e., the macroscopic framework completely collapses, and the damage variable reaches its extreme value of 1). This exponential decay term utilizes the mathematical asymptote property of the negative exponential function to forcibly set a ceiling for numerical growth at the bottom layer, perfectly matching the insurmountable nature of rock mass damage. This reflects the objective physical reality. Simultaneously, as the rock mass is continuously damaged, the remaining intact rock skeleton gradually depletes, leading to a diminishing marginal return on the additional damage caused by the same scale of energy increase in the later stages of evolution. This exponential decay term, through its gradually decreasing slope derivative characteristic, extremely accurately reproduces this dissipative passivation phenomenon of damage evolution in rock mechanics. Furthermore, when the value of this exponential decay term is small (i.e., the negative exponent approaches 1), it means that the energy accumulation rate and total amount brought about by the blasting disturbance are still within the system's safe capacity, and the rock mass has not yet induced an exponentially accelerated damage chain reaction. When the value of this exponential decay term is large and infinitely close to 1 (i.e., the negative exponent approaches 0), it means that the measured energy is explosively increasing, the damage evolution of the rock mass has entered the late accelerated stage, and the internal micro-cracks have fully connected to form a macroscopic fracture surface, exhibiting an irreversible rapid deterioration and collapse trend. This term utilizes the asymptotic convergence characteristics of the negative exponential function to profoundly anchor the mechanical convergence boundary of the late-stage energy surge but structural destruction tends towards the limit value.

[0043] By multiplying the fractional damping term and the exponential decay term corresponding to the acquisition point, the cumulative damage index of the acquisition point is calculated. Through nonlinear mapping and normalization constraints on the energy scale, not only is the entire process of rock mass damage evolution accurately characterized, but the evaluation error caused by the difference in electromagnetic background energy at different depth acquisition points is also effectively eliminated. The specific calculation formula is as follows: In the formula, To calculate the cumulative damage index for this sampling point, For fractional damping terms, This represents the cumulative marginal energy at the collection point. In this embodiment, the preset reference energy is set to the value of the undisturbed, intact rock mass. , This is an exponentially decaying term.

[0044] The cumulative damage index refers to the overall deterioration of deep rock masses after blasting disturbance, where the micro-fracture clusters evolve to a point where the macro-structural support is completely lost. This cumulative damage index is modulated nonlinearly by a fractional damping term and an exponential decay term. The fractional damping term dominates the elastic threshold suppression in the low-energy range, while the exponential decay term dominates the evolution limit calibration in the high-energy range. These two terms, through multiplication, form a dual-verification and mutually constraining coupling mechanism. In the early stages of blasting or when the rock mass integrity is high, the strong hysteresis effect of the damping term forces intervention, keeping the total damage index low and effectively avoiding false positives caused by slight vibrations in engineering practice. However, once the measured cumulative marginal energy exceeds the preset reference energy physical benchmark, the restriction of the fractional damping term is lifted, and the exponential decay term takes over, driving the cumulative damage index to rise rapidly and ultimately converge to a mathematically strictly constrained value. Within the dimensionless standard range, this solves the technical problem of the difficulty in applying traditional absolute energy threshold values ​​across different geological conditions and radar equipment. Furthermore, when the cumulative damage index is extremely small (approaching 0), it means that the blasting disturbance is completely within the elastic self-absorption range of the rock mass, the surrounding rock structure in the monitoring area is intact, and no special support measures are required; when the cumulative damage index increases sharply and approaches the extreme value of 1, it indicates that irreversible macroscopic through-damage has occurred inside the rock mass, and the supporting framework of the surrounding rock has substantially disintegrated. Thus, the calculation of the cumulative damage index breaks through the qualitative limitations of traditional rock mass stability evaluation that relies on manual experience or a single absolute threshold.

[0045] Step 5: Delineate the damaged area based on the spatial continuous distribution characteristics of the cumulative damage index exceeding the preset threshold at each collection point, and compare the maximum value of the cumulative damage index within the damaged area with the preset level interval to output the surrounding rock damage result.

[0046] In this embodiment, due to the complexity of the surrounding rock environment in deep tunnels, radar echoes often contain isolated numerical variations caused by localized natural discontinuous joints or random water seepage. If damage warnings are based solely on single-point values, misjudgments due to data jumps are highly likely. Therefore, this embodiment obtains the spatial coordinates of each acquisition point on the test line, compares the calculated cumulative damage index with a preset threshold, and filters out abnormal acquisition points exceeding the threshold. Since hard rock exhibits weak elastic deformation and acoustic emission oscillations in the initial unloading stage, its cumulative damage index often fluctuates around 0.1. This embodiment sets the preset threshold to 0.15, effectively filtering out natural background fluctuations and ensuring that only nodes with substantial damage are clustered.

[0047] By utilizing the constraints of spatial dimensions, discrete random noise unrelated to blasting disturbances is eliminated from a physical perspective. This embodiment clusters consecutively adjacent anomalous acquisition points in spatial location into the same set based on spatial coordinates, and delineates the physical spatial boundary corresponding to this set as an independent destruction region, ensuring that the identified target area possesses structural and continuous destruction characteristics, rather than isolated physical performance fluctuations.

[0048] In this embodiment, for any defined damaged area, the maximum cumulative damage index of all abnormal acquisition points within the corresponding set of the damaged area is extracted. By using the most severely damaged local layer within the area as a representative of the overall safety risk of the continuously damaged area, it is ensured that the assessment results are not masked by the mean calculation within the area. Subsequently, this maximum value is compared with a preset level interval.

[0049] The preset damage level range consists of progressively increasing intervals of mild, moderate, and severe damage. The cumulative damage index corresponds to... In the model evolution curve, the lower inflection point where the first derivative begins to increase nonlinearly is set as the boundary between the mild and moderate damage ranges, mapping the plastic yield point of the actual rock mass; simultaneously, the cumulative damage index is corresponding to... In the rock mass evolution curve, the upper inflection point where the second derivative changes from positive to negative is set as the boundary between the moderate and severe damage intervals, mapping the peak intensity point of the rock mass. Using a stepped quantitative classification, the abstract radar energy variation characteristics are deeply anchored to the physical stages of surrounding rock damage, thus transforming complex wavefield signals into intuitive engineering classifications. By dividing the rock into three independent intervals, the entire nonlinear degradation process of deep hard rock, from microfracture initiation to fracture convergence and then to a rapid loss of structural bearing capacity, can be accurately captured; effective interception of local random geological disturbances is achieved, and the health risk level of the surrounding rock is quantified.

[0050] When the maximum value falls within the minor damage range, it is determined that the damaged area has suffered minor damage in this blasting disturbance. When the maximum value falls within the moderate damage range, it is determined that the damaged area has suffered moderate damage in this blasting disturbance. When the maximum value falls within the severe damage range, the damaged area is determined to have suffered severe damage in this blasting disturbance.

[0051] In this embodiment, the range of mild damage is set as follows: The moderate damage range is set as The severely damaged area was set as Based on the measured statistical law that the damage index usually fluctuates around 0.1 due to the weak elastic deformation generated in the early stage of unloading of deep hard rock and the electromagnetic background noise, the preset threshold is set to 0.15 to effectively eliminate the background fluctuation and random interference in the undamaged state. Based on the physical transition point of the evolution of microcracks from isolated development to local convergence and connection in hard rock damage mechanics, the boundary between the light damage range and the moderate damage range is set to 0.40, which describes the critical value of moderate damage when microcracks evolve from isolated development to local convergence and connection. Based on the failure characteristics of the surrounding rock bearing skeleton in the critical fracture stage, which is characterized by a sharp drop in strength and a nonlinear surge in deformation rate, the boundary between the moderate damage range and the severe damage range is set to 0.75, which anchors the dangerous state of the rock mass structure losing its support force sharply and approaching the physical fracture limit. Based on the theory of continuous medium damage mechanics, the damage extreme value is set to 1.0, which means that the rock mass in the monitored area has completely left the elastic and plastic strengthening stage and entered the state of complete collapse. By setting a range, the microscopic and abstract radar reflected wave energy distortion index is objectively converged into a standardized risk criterion with direct engineering safety assessment and disaster prevention guidance value.

[0052] Table 1 is an example table of the spatial evolution of multidimensional physical characteristics and cumulative damage criteria of the surrounding rock of the roadway under test at 25 consecutive collection points.

[0053] Table 1: Comprehensive Data Table of Multidimensional Feature Parameters and Damage Evolution Table 1 shows the data demonstrating the accuracy of this method in quantifying and separating boundaries of different damage levels in the surrounding rock of a roadway under real-world testing conditions in a deep unloading environment. (Data collection points) and Although the disturbance amplitude remains at a high level ( However, due to the lack of high-frequency fracture sound signals, the frequency domain weighting term is extremely small, resulting in a low calculated cumulative marginal energy and a correspondingly low cumulative damage index. This reflects the conventional wavefield reflection characteristics of the surrounding rock in an undamaged state or under natural weak elastic oscillations, lacking substantial fracture evolution of the internal structure; sampling point and It exhibits a significant decrease in perturbation amplitude ( The rapidly increasing cumulative marginal energy, along with the cumulative damage index smoothly crossing the preset threshold of 0.15 and steadily rising, indicates that although microfractures are beginning to emerge within the surrounding rock, causing energy scattering, the numerical evolution exhibits a smooth progressive relationship due to the constraint of the fractional damping term, without any hypersensitive jumps; (collection point) The amplitude of the disturbance drops to its lowest point ( The accumulated marginal energy exhibits an unrestrained nonlinear surge, but at the same time, the accumulated damage index does not diverge infinitely under the convergence of the exponential decay term, but stabilizes in a high range. The data exceeded the critical threshold, indicating that radar reflected waves undergo significant scattering and severe energy loss in media experiencing substantial macroscopic fracturing. This table demonstrates that traditional monitoring methods based on a single amplitude threshold are prone to confusing local natural lithological fluctuations with actual blast damage, while this embodiment can accurately filter out false interference from the elastic background of the surrounding rock.

[0054] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0055] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented in software, the above embodiments can be implemented, in whole or in part, as a computer program product. Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution.

[0056] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0057] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A comprehensive monitoring and treatment method for the surrounding rock damage area in deep unloaded hard rock roadways under blasting disturbance, characterized in that, The specific steps include: For the acquisition points of deep unloaded hard rock roadways, radar reflection wave sequences of the surrounding rock before and after blasting disturbance are acquired within a preset time window. The two sequences are time-aligned and Hilbert transform is performed on them respectively, and they are reconstructed into the reference analytical sequence and disturbance analytical sequence of the acquisition point. The reference amplitude sequence and reference phase sequence of the acquisition point are extracted from the reference analytical sequence, and the disturbance amplitude sequence and disturbance phase sequence of the acquisition point are extracted from the disturbance analytical sequence. For each time node of the preset time window, an amplitude difference term is constructed based on the energy variation characteristics of the perturbation amplitude sequence and the reference amplitude sequence. A phase amplitude interference term is constructed by combining the deflection characteristics of the perturbation phase sequence and the reference phase sequence and the corresponding double amplitude characteristics. This term is nonlinearly coupled with the amplitude difference term and arranged in time sequence to obtain the time-shifted differential signal sequence of the acquisition point. Hilbert-Huang transform is performed on the time-shifted differential signal sequence to decompose the intrinsic mode function components. The time-varying amplitude and time-varying frequency of the target component at each time node are extracted. Based on the ratio characteristics of the time-varying frequency and the preset radar center frequency, the frequency domain weighting term of the corresponding node is constructed and multiplied by the square of the time-varying amplitude. The sequential summation is performed within the preset time window to obtain the cumulative marginal energy of the acquisition point. A fractional damping term is constructed based on the ratio of the square of the cumulative marginal energy to the preset reference energy. An exponential decay term is constructed based on the negative exponential decay characteristic of the ratio of the cumulative marginal energy to the preset reference energy. The fractional damping term is then multiplied with the fractional damping term to obtain the cumulative damage index at the sampling point. Based on the spatial continuous distribution characteristics of the cumulative damage index exceeding the preset threshold at each collection point, the damaged area is delineated, and the maximum value of the cumulative damage index within the damaged area is compared with the preset level range to output the surrounding rock damage result.

2. The method for comprehensive monitoring and treatment of surrounding rock damage areas in deep unloaded hard rock roadways under blasting disturbance as described in claim 1, characterized in that: The steps for defining the collection point and the preset time window specifically include: On the exposed surrounding rock surface of the deep unloading hard rock tunnel, longitudinal survey lines are planned along the axial direction of the tunnel, and transverse survey lines are planned along the cross-sectional contour of the tunnel. The detection grid line formed by the interweaving of the longitudinal survey lines and the transverse survey lines is defined as the line to be measured, and multiple collection points are set at equal intervals on the line to be measured. For any collection point on the line to be measured, determine the upper and lower boundaries of the preset monitoring depth of the target surrounding rock, and obtain the propagation speed of radar electromagnetic waves in the current surrounding rock mass; Based on the physical conversion relationship between depth and two-way propagation time, the upper and lower boundaries are converted into corresponding time start and time end points, and the interval between the time start and time end points is defined as a preset time window.

3. The method for comprehensive monitoring and treatment of surrounding rock damage areas in deep unloading hard rock roadways under blasting disturbance as described in claim 2, characterized in that: The steps of time-aligning the two sequences, reconstructing the analytical sequence, and extracting the amplitude and phase sequences specifically include: By extracting the starting point of the direct wave, time zero-point alignment is performed on the radar reflected wave sequences before and after the blasting disturbance; When performing the Hilbert transform, the radar reflection wave sequence before the blast disturbance, after time zero-point alignment, is mapped from the one-dimensional real number domain to the reference analytical sequence of the acquisition point on the complex plane, and the radar reflection wave sequence after the blast disturbance, after time zero-point alignment, is mapped from the one-dimensional real number domain to the disturbance analytical sequence of the acquisition point on the complex plane. Calculate the modulus sequence and complex argument sequence of the reference analytical sequence in the complex plane, and extract them as the reference amplitude sequence and reference phase sequence of the acquisition point, respectively. Calculate the modulus sequence and complex argument sequence of the perturbation analytical sequence in the complex plane, and extract them as the perturbation amplitude sequence and perturbation phase sequence of the acquisition point, respectively.

4. The method for comprehensive monitoring and treatment of surrounding rock damage areas in deep unloaded hard rock roadways under blasting disturbance as described in claim 1, characterized in that: The step of constructing the amplitude difference term based on the energy variation characteristics of the perturbation amplitude sequence and the reference amplitude sequence specifically includes: For any time point within the preset time window: Calculate the numerical difference between the perturbation amplitude sequence and the reference amplitude sequence at this time node and label it as the first difference. Calculate the square of the first difference and construct an amplitude difference term that characterizes the absolute change of reflected energy at this time node. Calculate the phase difference between the perturbation phase sequence and the reference phase sequence at this time node and label it as the second difference value. Calculate half of the second difference value and take its corresponding square sinusoidal value to construct the deflection feature at this time node. Calculate the product of the perturbation amplitude sequence and the reference amplitude sequence at this time node, and take four times the value of the product to construct the dual amplitude feature at this time node; By multiplying the dual-amplitude feature and the deflection feature at this time point, the phase-amplitude interference term characterizing waveform delay and distortion amplification is obtained. The amplitude difference term and phase interference term at this time node are summed, and the square root operation is performed on the summation result to achieve nonlinear coupling. The time-shift differential signal value corresponding to this time node is calculated. All time nodes within the preset time window are traversed, and the time-shift differential signal values ​​at all time nodes are extracted and combined in chronological order to form the time-shift differential signal sequence of this acquisition point. The specific calculation formula is as follows: In the formula, For the first The time-shifted differential signal values ​​calculated at each time point and The perturbation amplitude sequence and the reference amplitude sequence are respectively at the 1st... Values ​​at each time point and The perturbation phase sequence and the reference phase sequence are respectively at the 1st... Values ​​at each time point This refers to the time node number.

5. The method for comprehensive monitoring and treatment of surrounding rock damage areas in deep unloaded hard rock roadways under blasting disturbance as described in claim 1, characterized in that: The specific steps for obtaining the cumulative marginal energy at the collection point are as follows: The time-shifted differential signal sequence is characterized as a non-stationary signal bearing the characteristics of surrounding rock damage. The Hilbert-Huang transform is used to adaptively decompose it and map it to the instantaneous time-frequency domain to deconstruct the dynamic frequency mutation and energy fluctuation induced by rock mass fracture, thereby extracting the time-varying amplitude and time-varying frequency at each time node. For any given time point, calculate the ratio of the time-varying frequency to the preset radar center frequency at that time point. Summate this ratio with the value 1 and extract the natural logarithm of the sum to construct the frequency domain weighting term representing the nonlinear gain of high-frequency micro-damage corresponding to that time point. The instantaneous energy corresponding to the time node is obtained by multiplying the frequency domain weighting term corresponding to the time node with the square of the time-varying amplitude at the time node. The instantaneous energy value corresponding to each time node is multiplied by a preset sampling time interval, and the product results corresponding to all time nodes within the preset time window are sequentially summed to calculate the cumulative marginal energy of the sampling point. The specific calculation formula is as follows: In the formula, To calculate the cumulative marginal energy at this collection point, The total number of sampling points within the preset time window. The time node number is within the range of , For the first The time-varying amplitude corresponding to each time point To target time-shifted differential signals within a preset time window, the single dominant intrinsic mode function component with the highest total energy proportion within the entire time window, selected through empirical mode decomposition, is located at a specific time node. The time-varying frequency below, To preset the radar center frequency, This is the preset sampling time interval.

6. A comprehensive monitoring and treatment method for the surrounding rock damage area in deep unloading hard rock roadways under blasting disturbance as described in claim 1, characterized in that: The specific steps for obtaining the cumulative damage index of the sampling point are as follows: Extract the cumulative marginal energy at the collection point and calculate the square of the cumulative marginal energy; Calculate the square of the preset reference energy, and sum the square of the cumulative marginal energy with the square of the preset reference energy to obtain a first sum value; Calculate the ratio of the square of the accumulated marginal energy to the first sum, and construct a fractional damping term corresponding to the acquisition point to characterize the numerical suppression effect in the early stage of damage; Calculate the ratio of the accumulated marginal energy to the preset reference energy and calibrate it as the first ratio. Extract the decay value with the natural constant as the base and the opposite of the first ratio as the exponent. Calculate the difference between the value 1 and the decay value to construct the exponential decay term corresponding to the acquisition point that characterizes the late-stage convergence boundary of the damage. The cumulative damage index of the acquisition point is calculated by multiplying the fractional damping term and the exponential decay term corresponding to that acquisition point. The specific calculation formula is as follows: In the formula, To calculate the cumulative damage index for this sampling point, This represents the cumulative marginal energy at the collection point. This is the preset reference energy for undisturbed, intact rock masses.

7. A comprehensive monitoring and treatment method for the surrounding rock damage area in a deep unloading hard rock roadway under blasting disturbance, as described in claim 2, characterized in that: The specific planning steps for the damaged area include: The spatial coordinates of each sampling point on the line to be tested are obtained, and the cumulative damage index corresponding to each sampling point is compared with the preset threshold. Abnormal collection points that have a cumulative damage index exceeding the preset threshold are selected. Based on the spatial coordinates, consecutively adjacent abnormal collection points in space are clustered into the same set, and the physical spatial boundary corresponding to the set is defined as an independent damage area.

8. A comprehensive monitoring and treatment method for the surrounding rock damage area in a deep unloading hard rock roadway under blasting disturbance, as described in claim 7, characterized in that: The specific steps for outputting the surrounding rock damage results are as follows: For any damaged area, extract the maximum value of the cumulative damage index of each abnormal collection point in the set corresponding to that damaged area; The maximum value is compared with a preset level range, which consists of a progressively increasing range of mild damage, moderate damage, and severe damage. When the maximum value falls within the minor damage range, it is determined that the damaged area has suffered minor damage in this blasting disturbance. When the maximum value falls within the moderate damage range, it is determined that the damaged area has suffered moderate damage in this blasting disturbance. When the maximum value falls within the severe damage range, the damaged area is determined to have suffered severe damage in this blasting disturbance.

Images