An active waveguide acoustic emission monitoring method for monitoring the stability of a slope

Abstract

The application discloses an active waveguide acoustic emission monitoring method for slope stability monitoring and relates to the technical field of slope stability monitoring. The method comprises the following steps: drilling a hole in a slope monitoring area and placing a waveguide rod with an active excitation element in the hole, pouring slurry and solidifying the slurry to form a rigid integrated waveguide sensing structure which is acoustically coupled with a rock mass; through the rigid integrated waveguide sensing structure, synchronously collecting passive signals of rock mass micro-cracking and active reference signals triggered by the active excitation element, and obtaining mixed signals; taking a stable waveform inherent to the active reference signals in the rigid integrated waveguide sensing structure as a reference, separating a signal base from the synchronously collected signals, and extracting the passive signals after removing the signal base. The application realizes accurate separation of the signal base, effectively extracts weak rock mass cracking effective signals, and significantly improves the accuracy and reliability of signal identification in slope stability monitoring.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of slope stability monitoring, and particularly relates to an active waveguide acoustic emission monitoring method for slope stability monitoring. BACKGROUND

[0002] In the field of geotechnical engineering safety monitoring, acoustic emission technology is used to perceive the micro-fracture activity inside the slope rock mass to achieve stability early warning. The existing conventional method usually implants sensors or waveguide rods in the slope to evaluate by collecting and analyzing the intensity, frequency and other parameters of the acoustic emission signal.

[0003] However, in the actual complex geological environment, the coupling interface between the sensor and the rock mass, environmental vibration and the electronic noise of the monitoring system itself form a complex and time-varying signal base together, which causes the real and weak rock mass fracture signal to be seriously submerged.

[0004] Therefore, the existing technology generally lacks a reliable means to stably and accurately separate the effective fracture signal from such strong background interference, and the data analysis is mostly based on mixed signals that are not sufficiently purified, which fundamentally causes errors in subsequent parameter extraction and state evaluation, and seriously restricts the accuracy and early warning of the early warning. SUMMARY

[0005] The present application provides an active waveguide acoustic emission monitoring method for slope stability monitoring to solve the problems raised in the background.

[0006] To achieve the above-mentioned purpose, the present application provides an active waveguide acoustic emission monitoring method for slope stability monitoring, comprising:

[0007] S1, drilling a hole in the slope monitoring area and placing a waveguide rod with an active excitation element, pouring grout to solidify, and forming a rigid integrated waveguide sensing structure acoustically coupled with the rock mass;

[0008] S2, synchronously collecting the passive signal of the rock mass micro-fracture and the active reference signal triggered by the active excitation element through the rigid integrated waveguide sensing structure to obtain a mixed signal;

[0009] S3, taking the stable waveform inherent in the active reference signal in the rigid integrated waveguide sensing structure as a reference to separate the signal base from the synchronously collected signal, and after removing the signal base, extracting the passive signal;

[0010] S4, extracting the fracture frequency band parameters modulated by the conduction characteristics of the rigid integrated waveguide sensing structure from the passive signal; at the same time, extracting the propagation time sequence parameters and amplitude parameters representing the physical state change of the rigid integrated waveguide sensing structure from the active reference signal, which together constitute the active propagation state parameters;

[0011] S5. Jointly analyze the rupture frequency band parameters and the active propagation state parameters, establish their change correlation, and generate comprehensive state parameters;

[0012] S6. Based on the temporal evolution pattern of the comprehensive state parameters, determine the internal damage development of the rock mass and the stress state perceived by the rigid coupling body, and generate early warning information.

[0013] Preferably, the step of separating the signal substrate from the synchronously acquired signal based on the inherent stable waveform of the active reference signal in the rigid integrated waveguide sensing structure includes:

[0014] Acquire the waveform of the active reference signal within a preset stabilization period and establish a reference waveform template;

[0015] Perform waveform matching operation between the mixed signal and the reference waveform template to locate the part of the mixed signal whose shape similarity to the reference waveform template meets the preset conditions;

[0016] The portion whose morphological similarity meets the preset conditions is removed from the mixed signal to obtain the residual signal after substrate removal.

[0017] Preferably, the step of extracting the passive signal after removing the signal substrate includes:

[0018] Time-domain energy peak detection is performed on the residual signal to identify all transient pulse events that exceed the energy threshold;

[0019] The time-frequency distribution characteristics of transient pulse events are compared with a pre-defined rupture signal feature library, and pulse events with matching characteristics are selected as passive signals.

[0020] Preferably, the extraction of the broken frequency band parameter modulated by the conduction characteristics of the rigid integrated waveguide sensing structure from the passive signal includes:

[0021] Perform spectral transformation on each event in the passive signal to obtain the spectrum of each event;

[0022] From the spectrum of each event, locate the continuous frequency band with the highest energy concentration, and determine the center position and bandwidth parameters of the frequency band;

[0023] Arrange the frequency band center position and width parameters corresponding to each event in chronological order to form a sequence of frequency band break parameters.

[0024] Preferably, the extraction of propagation timing parameters and amplitude parameters characterizing the physical state changes of the rigid integrated waveguide sensing structure itself from the active reference signal includes:

[0025] Identify the start time of each active reference signal event and determine its propagation time in the rigid integrated waveguide sensing structure;

[0026] Measure the peak amplitude of each active reference signal event;

[0027] The propagation timing variation parameter is determined by the difference in propagation duration between adjacent events; the amplitude variation parameter is determined by the ratio of the peak amplitudes of adjacent events.

[0028] Preferably, the two together constitute the active propagation state parameter, including:

[0029] Align the propagation timing variation parameter with the amplitude variation parameter using the same time coordinate;

[0030] The propagation time sequence change value and amplitude change value corresponding to each moment are combined into a state vector;

[0031] Arrange the state vectors at all times in chronological order to form an active propagation state parameter sequence.

[0032] Preferably, the step of jointly analyzing the rupture frequency band parameters and the active propagation state parameters to establish their change correlation includes:

[0033] Synchronize the rupture frequency band parameter sequence and the active propagation state parameter sequence on the time axis;

[0034] Identify and compare the points of change in the rupture frequency band parameters with the points of change in the active propagation state parameters;

[0035] The time intervals in which the rupture frequency band parameter and the active propagation state parameter both change significantly are marked as strong correlation periods.

[0036] Preferably, the generation of comprehensive state parameters includes:

[0037] During periods of strong correlation, preset weights are assigned to the intensity of changes in the rupture frequency band parameter and the intensity of changes in the active propagation state parameter, respectively.

[0038] The two change intensity values, after being assigned preset weights, are superimposed to obtain the comprehensive state parameter value for that period.

[0039] The comprehensive state parameter values ​​for all time periods are concatenated in chronological order to form a complete comprehensive state parameter sequence.

[0040] Preferably, determining the internal damage development of the rock mass based on the temporal evolution pattern of the comprehensive state parameters includes:

[0041] The integrated state parameter sequence is divided into multiple consecutive analysis windows;

[0042] Within each analysis window, extract the trend and fluctuation characteristics of the comprehensive state parameter values;

[0043] The trend and fluctuation characteristics are matched with a pre-stored damage evolution pattern library to determine the most matching damage development stage.

[0044] Preferably, generating the early warning information includes:

[0045] The corresponding warning level is determined by comparing the damage development stage with the preset warning level mapping relationship;

[0046] The warning level is combined with the current stage of damage development and key strongly correlated time periods to form the warning information.

[0047] Compared with the prior art, the present invention has the following beneficial effects:

[0048] 1. By constructing a rigid integrated waveguide sensing structure that is acoustically coupled with the rock mass, passive signals and active reference signals of micro-fractures in the rock mass are acquired simultaneously. The inherent stable waveform of the active reference signal is used as a reference to achieve accurate separation of the signal substrate, effectively extracting weak effective signals of rock mass fractures, significantly improving the accuracy and reliability of signal identification in slope stability monitoring, and providing solid technical support for the early perception of internal damage in the rock mass.

[0049] 2. By refining the design of signal separation logic, feature parameter extraction methods, parameter joint analysis modes, and time series evolution judgment rules, the comprehensive state parameters are further enhanced to characterize rock mass damage and stress state, making the judgment of damage development stages more accurate and timely, improving the comprehensiveness and practical value of early warning information, and ensuring the efficient advancement of slope safety monitoring work. Attached Figure Description

[0050] Figure 1 This is a flowchart illustrating an active waveguide acoustic emission monitoring method for slope stability monitoring, provided in an embodiment of the present invention.

[0051] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0052] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0053] This application provides an active waveguide acoustic emission monitoring method for slope stability monitoring. The executing entity of this active waveguide acoustic emission monitoring method for slope stability monitoring includes, but is not limited to, at least one of the following electronic devices that can be configured to execute the method provided in this application embodiment: a server, a terminal, etc. In other words, an active waveguide acoustic emission monitoring method for slope stability monitoring can be executed by software or hardware installed on a terminal device or a server device. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster. The server can be an independent server or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDN), and big data and artificial intelligence platforms.

[0054] Example 1, referring to Figure 1 The diagram shown is a flowchart illustrating an active waveguide acoustic emission monitoring method for slope stability monitoring according to an embodiment of the present invention. In this embodiment, the active waveguide acoustic emission monitoring method for slope stability monitoring includes:

[0055] S1. Drill holes in the slope monitoring area and insert waveguide rods with active excitation elements, then inject grout to solidify and form a rigid integrated waveguide sensing structure that is acoustically coupled with the rock mass.

[0056] Specifically, multiple monitoring points are selected in the slope monitoring area. Up to three monitoring points can be selected, with a spacing of 10m between them. Each monitoring point is drilled vertically using a drilling machine to a depth of 4m, ensuring that the waveguide rod is exposed 0.5m and buried 3.5m into the rock mass, with a hole diameter of 50mm.

[0057] Slowly insert the waveguide rod with the built-in piezoelectric ceramic transducer into the borehole, ensuring that the transducer is located in the middle section of the waveguide rod, generally 1.5m from the bottom of the hole, corresponding to the stress concentration sensitive area of ​​the rock mass;

[0058] Furthermore, the prepared cement-based grout is injected into the borehole using pressure grouting to ensure that the grout fills the gap between the waveguide rod and the borehole wall without any air bubbles remaining.

[0059] Furthermore, after curing for a certain period of time, generally 72 hours, at an ambient temperature of 25℃ and a humidity of ≥80%, the waveguide rod, the cured slurry, and the surrounding granite body form a rigid integrated waveguide sensing structure with continuous acoustic impedance.

[0060] S2. Through a rigid integrated waveguide sensing structure, the passive signal of rock micro-fracture and the active reference signal triggered by the active excitation element are simultaneously acquired to obtain a mixed signal.

[0061] Specifically, the top of the waveguide rod of the rigid integrated waveguide sensing structure is rigidly connected to the signal input terminal of the data acquisition instrument, the active excitation element is connected to the output terminal of the signal generator through a wire, and the data acquisition instrument and the signal generator are connected through a synchronous trigger line to achieve timing synchronization.

[0062] Specifically, based on the inherent operating frequency of the active excitation element to ensure signal transmission efficiency, the signal generator presets single-pulse excitation parameters: frequency 300kHz, amplitude 5V. In order to balance the integrity of the acquisition and energy consumption, the control trigger period can be set to 100ms.

[0063] To cover the frequency range of passive signals of rock microfractures (10kHz-1MHz) and active reference signals (300kHz), the data acquisition instrument was preset to a sampling frequency of 2MHz.

[0064] Specifically, the sampling duration is consistent with the triggering cycle. After the synchronization mode is turned on, the signal generator triggers the active excitation element to generate an active reference signal. The passive signal of rock micro-fracture is synchronously transmitted through the waveguide rod. The two signals enter the same acquisition channel and are naturally superimposed to obtain a mixed signal.

[0065] S3. Using the inherent stable waveform of the active reference signal in the rigid integrated waveguide sensing structure as a reference, the signal substrate is separated from the synchronously acquired signal, and the passive signal is extracted after the signal substrate is removed.

[0066] In this embodiment, the signal substrate is separated from the synchronously acquired signal based on the inherent stable waveform of the active reference signal in the rigid integrated waveguide sensing structure, including:

[0067] Acquire the waveform of the active reference signal within a preset stabilization period and establish a reference waveform template;

[0068] Perform waveform matching operation between the mixed signal and the reference waveform template to locate the part of the mixed signal whose shape similarity to the reference waveform template meets the preset conditions;

[0069] The portion whose morphological similarity meets the preset conditions is removed from the mixed signal to obtain the residual signal after substrate removal.

[0070] Similarity is determined by the following methods to meet the preset conditions:

[0071] Set similarity threshold When morphological similarity Not less than When the preset conditions are met, it is determined that the conditions are met.

[0072] Morphological similarity This is obtained as follows: Let the discrete sequence of the reference waveform template be... The discrete sequence of the matching window in the mixed signal is ;

[0073] Calculate sequence and Normalized cross-correlation coefficients And this coefficient is used as the morphological similarity. ,Right now:

[0074] ;

[0075] in, For sequence length, For sequence The mean, For sequence The mean.

[0076] Specifically, the preset stabilization period for acquiring the active reference signal is set to 10 trigger cycles. This period is derived from the trigger cycle of the active excitation element and is generally in the range of 5-20 trigger cycles. For example, it is set to 10 trigger cycles. Active reference signal waveforms that are not disturbed by rock micro-fractures within these 10 cycles are collected. All waveforms are superimposed one by one according to the sampling points and the arithmetic mean is taken to obtain a smooth and noise-free reference waveform template. This operation can enhance the inherent stability characteristics of the active reference signal and provide a precise reference for subsequent waveform matching.

[0077] Specifically, the length of the window to be matched is related to the sequence length of the reference waveform template. To maintain consistency, based on a 2MHz sampling frequency and a 300kHz active reference signal frequency, it is possible to... With 2000 sampling points, this length can fully cover 3 signal cycles, ensuring the integrity of the waveform shape.

[0078] Specifically, using a sliding step of 10 sampling points, the mixed signal is segmented from the starting position and... Matching windows of uniform length are used, and each window is compared with the reference waveform template in terms of shape. The sliding step size is set to 10 sampling points, which can ensure matching accuracy while improving processing efficiency and avoiding redundant calculations caused by too small a step size.

[0079] Specifically, the sequence The discrete sampled values ​​of the reference waveform template are obtained by decomposing the average waveform within the stable period, and the sequence is... The discrete sampled values ​​of the mixed signal to be matched window are obtained by segmenting the mixed signal piece by piece, and the sequence is calculated. Each sampling point and sequence The difference of the arithmetic means of all sampling points is used to calculate the sequence similarly. Each sampling point and sequence The difference between the arithmetic means of all sampling points is calculated by multiplying the two sets of differences one by one at corresponding positions, summing the results, and then dividing by the sequence. The sum of squares of all differences and the sequence The square root of the product of the sum of the squares of all differences is the morphological similarity score. This calculation method can quantify the degree of morphological fit between two sets of waveforms; the closer the value is to 1, the more consistent the morphology.

[0080] Specifically, considering the transmission characteristics of the active reference signal in the rigid integrated waveguide sensing structure, its similarity to the template is typically higher than 0.85, while the similarity between the passive signal of rock microfracture and the template is generally lower than 0.7. Therefore, a similarity threshold can be set... Setting the threshold to 0.85 allows for precise differentiation between active and passive reference signals, comparing the morphological similarity of each window to be matched. and ,when If the value is not less than 0.85, the window is considered a signal base.

[0081] Specifically, all windows identified as signal substrates are removed segment by segment from the mixed signal. During removal, the unmatched parts between windows are retained, resulting in the residual signal after substrate removal. This operation completely separates the signal substrate by accurately locating and removing the components belonging to the active reference signal in the mixed signal.

[0082] In summary, the rigid integrated waveguide sensing structure ensures the stability of the active reference signal, resulting in more thorough base separation. The residual signal contains only passive signals from micro-fractures in the rock mass, thus enabling accurate extraction of passive signals. This extraction method avoids the passive signal distortion caused by signal interference in traditional methods, improves the signal-to-noise ratio of passive signals, and provides accurate data support for subsequent slope stability analysis.

[0083] In this embodiment of the invention, after removing the signal substrate, the passive signal is extracted, including:

[0084] Time-domain energy peak detection is performed on the residual signal to identify all transient pulse events that exceed the energy threshold;

[0085] The time-frequency distribution characteristics of transient pulse events are compared with a pre-defined rupture signal feature library, and pulse events with matching characteristics are selected as passive signals.

[0086] Specifically, based on the initial stage after the rigid integrated waveguide sensing structure is installed and no rock micro-fractures have occurred, the background noise signal is continuously collected for 1 hour. The arithmetic mean of the squares of the amplitudes of all sampling points during this period is calculated to obtain the average energy of the background noise. This acquisition method can truly reflect the inherent noise level of the monitoring environment.

[0087] Specifically, the average energy of background noise is usually calibrated to 0.01V² in actual monitoring scenarios, and the energy threshold is generally 3-5 times the average energy of background noise. For example, the energy threshold parameter is set to 0.04V², so as to ensure that the weak pulse signal generated by the micro-fracture of the rock mass can be captured, while eliminating low-energy noise generated by environmental vibration and other interference, thus avoiding missed detection or false detection.

[0088] Furthermore, based on this energy threshold setting, a fixed-length sliding window matching the period of the active reference signal is used to traverse the residual signal. The sliding window length is set to 3 active reference signal periods. The energy value of each window is compared with the energy threshold of 0.04V². The signal segment corresponding to the window whose energy value exceeds the threshold is the transient pulse event.

[0089] Specifically, in order to cover the micro-fracture signal characteristics of common rock mass types on slopes, micro-fracture signals of typical lithologies such as granite, sandstone, and shale under different stress levels were collected. Time-frequency analysis was performed on these signals, and the core time-frequency distribution characteristics were extracted, including a frequency range of 10kHz-500kHz, an amplitude that shows a rapid increase followed by a slow decrease, and a signal duration of 10μs-100μs. These characteristics were then organized and archived to form a pre-set fracture signal feature library.

[0090] After establishing the feature library based on this, each identified transient pulse event is decomposed to obtain its frequency-time variation curve and amplitude-time variation curve. The frequency range, amplitude variation trend, and duration of the event are extracted as features to be compared. The features to be compared are then compared one by one with the standard features in the rupture signal feature library.

[0091] Specifically, the preset feature matching threshold is 80%. If the feature matching degree reaches 80% or more, it is determined to be a signal generated by micro-fracture of rock mass. This comparison method can accurately eliminate false pulse events, and the pulse events with feature matching that are finally selected are passive signals.

[0092] S4. Extract the break band parameters modulated by the conduction characteristics of the rigid integrated waveguide sensing structure from the passive signal; at the same time, extract the propagation timing parameters and amplitude parameters characterizing the physical state changes of the rigid integrated waveguide sensing structure itself from the active reference signal. The two together constitute the active propagation state parameters.

[0093] In this embodiment of the invention, extracting the broken frequency band parameter modulated by the conduction characteristics of the rigid integrated waveguide sensing structure from the passive signal includes:

[0094] Perform spectral transformation on each event in the passive signal to obtain the spectrum of each event;

[0095] From the spectrum of each event, locate the continuous frequency band with the highest energy concentration, and determine the center position and bandwidth parameters of the frequency band;

[0096] Arrange the frequency band center position and width parameters corresponding to each event in chronological order to form a sequence of rupture frequency band parameters;

[0097] The continuous frequency band with the highest energy concentration is determined in the following way:

[0098] For the spectrum of each event, its frequency-amplitude sequence is calculated. , Normalization is performed to obtain the normalized amplitude. ;

[0099] Define arbitrary continuous frequency bands (in Energy concentration This is the ratio of the integral of the normalized amplitude within this frequency band to the integral of the normalized amplitude over the entire frequency band, i.e.:

[0100] ;

[0101] in, Frequency intervals; by traversing all possible consecutive frequency bands , search for The largest frequency band, as the continuous frequency band with the highest energy concentration.

[0102] Specifically, to accurately reflect the frequency characteristics of passive signals after transmission through the rigid integrated waveguide sensing structure, the spectral transformation needs to be performed based on a fixed number of sampling points. The number of sampling points is derived from the sampling frequency and the signal duration. The sampling frequency is preset to 2MHz, and the signal duration corresponds to 10μs-100μs of transient pulse events. Setting the number of sampling points to 200 can completely cover the signal frequency range while ensuring spectral resolution and frequency spacing. The frequency interval is set to 1kHz. For each event in the passive signal, a complete signal segment is extracted according to the preset number of sampling points. The time-domain signal is converted into a sequence corresponding to frequency and amplitude through spectrum transformation, and the spectrum of each event is obtained.

[0103] Specifically, in order to eliminate the influence of the difference in amplitude of different passive signal events on energy judgment, it is necessary to first normalize the frequency amplitude sequence of each event. The normalized amplitude is obtained by dividing the amplitude corresponding to each frequency by the largest amplitude in the spectrum.

[0104] It should be noted that this method can unify all amplitudes to the 0-1 range, ensuring consistency in energy comparison and frequency spacing. Using the 1kHz frequency from the spectrum transformation, first calculate the sum of the products of all normalized amplitudes and frequency intervals within each continuous frequency band, then divide this sum by the sum of the products of all normalized amplitudes and frequency intervals within the entire frequency band to obtain the energy concentration of that frequency band.

[0105] In summary, this embodiment can intuitively reflect the relative enrichment of energy within a frequency band. The larger the value, the more concentrated the energy. In order from low frequency to high frequency, all possible continuous frequency bands are selected in sequence, and the energy concentration of each frequency band is calculated one by one. The frequency band with the largest energy concentration is selected as the continuous frequency band with the highest energy concentration. Then, the center position and width are calculated through the start frequency and end frequency of the frequency band. The center position is the arithmetic mean of the start frequency and the end frequency, and the width is the end frequency minus the start frequency.

[0106] Specifically, the development of microfractures in rock mass has a clear temporal sequence. Arranging parameters in chronological order can fully present the fracture evolution process, providing temporal data support for slope stability analysis. The frequency band center position and width parameters corresponding to each passive signal event are recorded sequentially according to the time sequence of their occurrence during the monitoring process, forming an ordered sequence of fracture frequency band parameters.

[0107] In summary, the fracture frequency band parameter sequence can reflect the modulation effect of the transmission characteristics of the rigid integrated waveguide sensing structure on passive signals. This is because the material and structure of the waveguide rod will selectively transmit signals of different frequencies. The frequency band with the highest energy concentration is the most effective fracture signal frequency band after modulation. The temporal changes of the sequence can accurately capture the development trend of micro fractures in rock mass, thereby improving the targeting and accuracy of monitoring.

[0108] In this embodiment of the invention, the propagation timing parameters and amplitude parameters characterizing the physical state changes of the rigid integrated waveguide sensing structure itself are extracted from the active reference signal, including:

[0109] Identify the start time of each active reference signal event and determine its propagation time in the rigid integrated waveguide sensing structure;

[0110] Measure the peak amplitude of each active reference signal event;

[0111] The propagation timing variation parameter is determined by the difference in propagation duration between adjacent events; the amplitude variation parameter is determined by the ratio of the peak amplitudes of adjacent events.

[0112] Specifically, in order to accurately locate the start and end points of the propagation of the active reference signal, the trigger threshold needs to be set based on the background noise amplitude. The background noise amplitude is obtained by statistical analysis of the signal during the initial stabilization phase of the sensing structure, and is generally 2-4 times the average amplitude of the background noise. The background noise amplitude is set to 0.5V, and the data acquisition instrument collects the signal at the top of the waveguide rod in real time at a preset sampling frequency of 2MHz. When the amplitude of the acquired signal first exceeds the 0.5V threshold, it is immediately recorded as the receiving time.

[0113] It should be noted that the propagation duration is determined by the time difference between the receiving time and the transmitting time. The acoustic conduction characteristics of the rigid integrated waveguide sensing structure are stable. Changes in the physical state of the structure, such as slight deformation, will cause changes in the signal propagation path, which in turn will cause changes in the propagation duration. Therefore, the propagation duration can directly characterize the physical state of the structure itself.

[0114] Specifically, to ensure the integrity of peak amplitude measurement, the measurement window needs to cover the complete propagation period of the active reference signal. The window length is derived based on the frequency and trigger period of the active reference signal, and is generally in the range of 100μs-300μs. The window length is set to 200μs because the period of the 300kHz active reference signal is about 3.3μs, and 200μs can cover 60 signal periods, which can completely include the entire process of the signal from rise to decay, avoiding the omission of peak values.

[0115] Specifically, within the signal segment corresponding to the determined propagation duration, the amplitude data of all sampling points are extracted one by one in the sampling order. All amplitude data are compared, and the amplitude with the largest value is selected as the peak amplitude of the active reference signal event. The degree of amplitude attenuation of the active reference signal when it propagates in the sensing structure is related to the physical state of the structure. Loosening or deformation of the structure will aggravate the amplitude attenuation. Therefore, the peak amplitude can effectively reflect the changes in the physical state of the structure itself.

[0116] Specifically, in order to capture the temporal change trend of the physical state of the structure, adjacent events are selected continuously in the triggering order of the active reference signal. The triggering period has been preset to 100ms, so the time interval between adjacent events is fixed at 100ms.

[0117] Specifically, the propagation timing change parameter is obtained by subtracting the propagation duration of the previous event from the propagation duration of the subsequent active reference signal event, and the amplitude change parameter is obtained by dividing the peak amplitude of the previous event by the peak amplitude of the subsequent event.

[0118] In summary, this extraction method can amplify minute changes in the physical state of the structure, such as slight increases in propagation time or slight decreases in amplitude caused by minor structural deformation. These changes can be accurately captured through difference and ratio calculations, solving the problem that it is difficult to detect minute changes by directly measuring single event parameters. The active propagation state parameters, together with the rupture frequency band parameters, can comprehensively characterize the physical state of the sensing structure from both propagation timing and amplitude dimensions, avoiding the limitations of single parameter monitoring and improving the comprehensiveness and accuracy of slope stability monitoring.

[0119] In this embodiment of the invention, the two together constitute the active propagation state parameter, including:

[0120] Align the propagation timing variation parameter with the amplitude variation parameter using the same time coordinate;

[0121] The propagation time sequence change value and amplitude change value corresponding to each moment are combined into a state vector;

[0122] Arrange the state vectors at all times in chronological order to form an active propagation state parameter sequence.

[0123] Specifically, to ensure that the propagation timing change parameter and the amplitude change parameter can be accurately correlated with the state of the rigid integrated waveguide sensing structure at the same moment, the time coordinate needs to be based on the triggering time of the active reference signal, and the timestamp accuracy is derived and set by the sampling frequency. The sampling frequency is 2MHz, corresponding to an interval of 0.5μs for each sampling point, and the timestamp can be set to 1μs.

[0124] Specifically, the trigger time of the next event in the adjacent active reference signal events corresponding to each propagation timing change parameter is extracted, and this time is used as the time coordinate of the propagation timing change parameter. The trigger time of the same next event corresponding to each amplitude change parameter is extracted in the same way and used as its time coordinate. The propagation timing change parameters with completely consistent time coordinates are matched one-to-one with the amplitude change parameters to complete the alignment process of the two according to the same time coordinate.

[0125] Specifically, in order to achieve multi-dimensional representation of the structural state at a single moment and avoid the limitation that a single parameter cannot fully reflect the state, for each aligned time coordinate, the propagation time sequence change value and amplitude change value corresponding to that moment are associated in a preset order. The preset order is fixed so that the propagation time sequence change value comes first and the amplitude change value comes last, thus facilitating subsequent unified analysis and forming a state vector containing two feature values. Each state vector fully carries the physical state information of the sensing structure at the corresponding moment in both the propagation time sequence and amplitude dimensions.

[0126] Specifically, in order to present the temporal evolution of the physical state of the rigid integrated waveguide sensing structure and provide basic data for the dynamic analysis of slope stability, all state vectors are arranged in chronological order according to their time coordinates to form an active propagation state parameter sequence.

[0127] In summary, this embodiment can intuitively reflect the trend of state vector change over time. Through this temporal combination, the structural propagation path change reflected by the propagation time change parameter and the energy attenuation change reflected by the amplitude change parameter complement each other. Even if there is slight structural loosening or deformation, it can be captured by the continuous fluctuation or abrupt change of the state vector in the sequence. This solves the problem that traditional single parameter monitoring is prone to missing slight state changes and greatly improves the accuracy and comprehensiveness of monitoring the physical state changes of the sensing structure itself.

[0128] S5. Jointly analyze the rupture frequency band parameters and the active propagation state parameters, establish their change correlation, and generate comprehensive state parameters.

[0129] In this embodiment, the breakage frequency band parameter and the active propagation state parameter are jointly analyzed to establish their change correlation, including:

[0130] Synchronize the rupture frequency band parameter sequence and the active propagation state parameter sequence on the time axis;

[0131] Identify and compare the points of change in the rupture frequency band parameters with the points of change in the active propagation state parameters;

[0132] The time intervals in which the rupture frequency band parameter and the active propagation state parameter both change significantly are marked as strong correlation periods.

[0133] Specifically, the occurrence time of the passive signal event corresponding to each parameter in the rupture frequency band parameter sequence is extracted, and the trigger time of the active reference signal corresponding to each state vector in the active propagation state parameter sequence is extracted. The time coordinates of both sequences are converted into a unified absolute time format. For parameters with slight deviations in time coordinates, linear interpolation of parameters at adjacent times is used to supplement them, so that each time point corresponds to both the rupture frequency band parameter and the active propagation state parameter, thus completing the synchronization processing of the two on the time axis.

[0134] It should be noted that, in order to accurately distinguish between normal fluctuations and significant changes in parameters, the threshold for significant changes needs to be determined based on the initial stable state of the parameters. The threshold for significant changes in parameters of the rupture frequency band is derived from the standard deviation of the parameters in the initial stable stage, and is generally in the range of 2-4 times the standard deviation. The threshold for significant changes can be set to 3 times the standard deviation.

[0135] It should also be noted that the threshold for significant change of the propagation time sequence parameter in the active propagation state parameters is also set to 3 times the initial standard deviation, and the threshold for significant change of the amplitude parameter is 0.1. Generally, when the amplitude change value is above 0.1, it indicates that the energy transfer efficiency of the structure has changed significantly, which is considered a significant change.

[0136] Specifically, the parameter value at each moment in the rupture frequency band parameter sequence is compared with the difference at the previous moment. If the difference exceeds the corresponding threshold, it is marked as a rupture frequency band change point. Active change points in the active propagation state parameter sequence are identified in the same way, and the time coordinates of all rupture frequency band change points and active change points are compared one by one.

[0137] It should be noted that, in order to accommodate the small time difference of parameter changes, the correlation time window is derived based on the signal propagation delay and acquisition delay, and generally ranges from 300ms to 700ms.

[0138] Specifically, the correlation time window is set to 500ms. All the comparison change point pairs are traversed. If the time difference between the change point of the broken frequency band and the active change point is within 500ms, the start time to the end time corresponding to the time difference is marked as a strongly correlated period.

[0139] In summary, the labeling method of this scheme can accurately pinpoint the causal relationship between micro-fractures in the rock mass and changes in the physical state of the sensing structure. By combining analysis, it eliminates the interference of misjudgment of a single parameter, such as environmental interference of passive signals or accidental fluctuations of active parameters. Only when both change significantly at the same time is it determined to be a valid correlation, ensuring that the strongly correlated period can truly reflect the coordinated changes of the slope rock mass and the sensing structure, and providing accurate core data support for the subsequent generation of comprehensive state parameters.

[0140] In this embodiment, the generation of comprehensive state parameters includes:

[0141] During periods of strong correlation, preset weights are assigned to the intensity of changes in the rupture frequency band parameter and the intensity of changes in the active propagation state parameter, respectively.

[0142] The two change intensity values, after being assigned preset weights, are superimposed to obtain the comprehensive state parameter value for that period.

[0143] The comprehensive state parameter values ​​for all time periods are concatenated in chronological order to form a complete comprehensive state parameter sequence.

[0144] Specifically, since the fracture frequency band parameter directly reflects the intensity of micro-fractures in the rock mass and the active propagation state parameter characterizes the changes in the physical state of the sensing structure, both have equal core value for slope stability monitoring. In order to balance the influence of the two parameters and avoid monitoring bias caused by the dominance of a single parameter, the preset weight is set to 0.5.

[0145] Specifically, by synchronously superimposing the two sets of correlated values, the comprehensive state parameter value within the strongly correlated period is obtained. This superposition method can effectively integrate the key information of rock mass fracture and structural state changes, avoiding misjudgment of state caused by incomplete information of a single parameter.

[0146] Specifically, to ensure that the comprehensive state parameter sequence covers the entire monitoring period, the comprehensive state parameter values ​​for non-strongly correlated periods are set according to preset benchmark values, which are derived from parameter data in the initial stable phase.

[0147] Specifically, the comprehensive state parameter value of the non-strongly correlated period is set to 0.1 according to the preset benchmark value. The comprehensive state parameter value of the strongly correlated period and the benchmark value of 0.1 of the non-strongly correlated period are arranged and connected in sequence according to the time axis to form a complete comprehensive state parameter sequence. This sequence can intuitively present the temporal evolution process of the slope from stable to abnormal changes.

[0148] In summary, by combining weighted superposition and time-series correlation, the comprehensive state parameter sequence not only preserves the coordinated change characteristics of rock mass fracture and structural state, but also highlights the abnormal information of strongly correlated periods through benchmark values. This solves the problem of scattered multi-parameter information and difficulty in comprehensive judgment in traditional monitoring, greatly improves the accuracy and practicality of slope stability monitoring, and provides direct data support for subsequent risk warning.

[0149] S6. Based on the temporal evolution pattern of the comprehensive state parameters, determine the internal damage development of the rock mass and the stress state perceived by the rigid coupling body, and generate early warning information.

[0150] In this embodiment of the invention, determining the internal damage development of the rock mass based on the temporal evolution pattern of comprehensive state parameters includes:

[0151] The integrated state parameter sequence is divided into multiple consecutive analysis windows;

[0152] Within each analysis window, extract the trend and fluctuation characteristics of the comprehensive state parameter values;

[0153] The trend and fluctuation characteristics are matched with a pre-stored damage evolution pattern library to determine the most matching damage development stage.

[0154] Identify the best-matching stage of injury development, including:

[0155] For the first in the pre-stored damage evolution pattern library There are several patterns, and their feature vectors are denoted as . ,in To reference trend direction, For reference fluctuation range value;

[0156] Let the feature vector extracted within the current analysis window be... ;

[0157] Calculate the current feature vector With each reference eigenvector Weighted Euclidean distance :

[0158] ;

[0159] in, and These are preset weighting coefficients for trend direction and fluctuation range, respectively. ; will be compared with the current feature vector Weighted Euclidean distance Minimal reference mode The corresponding stage of damage development is determined as the most suitable stage of damage development.

[0160] Specifically, the window length is derived based on the active reference signal triggering period and the time scale of rock mass damage evolution. The active reference signal triggering period is preset to 100 milliseconds. The characteristic evolution of rock mass microfractures needs to cover multiple parameter change points. The window length is generally in the range of 5-15 triggering periods. Setting the window length to 10 triggering periods, i.e. 1 second, can fully include multiple continuous comprehensive state parameter change points, avoiding the omission of key features due to an excessively short window or the blurring of features due to an excessively long window. The sliding step size is consistent with the window length, which is 1 second.

[0161] Specifically, starting from the beginning of the integrated state parameter sequence in chronological order, parameter data of 1 second duration are extracted as an analysis window until the end of the sequence, resulting in multiple consecutive analysis windows.

[0162] Specifically, within each analysis window, all comprehensive state parameter values ​​are extracted in chronological order. A linear fitting operation is performed on these parameter values, and the slope of the fitted line represents the trend direction. A slope of 0 represents a stable trend, a positive slope represents an upward trend, and a negative slope represents a downward trend.

[0163] In general, linear fitting can intuitively reflect the overall direction of parameter change over time. The fluctuation range is obtained by extracting the standard deviation of all comprehensive state parameter values ​​within the window. The standard deviation can accurately quantify the dispersion of parameter values, i.e. the magnitude of fluctuation. For example, if the slope of the parameter value after linear fitting within a certain window is 0.4, then the trend is upward. If the extracted standard deviation is 0.2, then the fluctuation range is 0.2.

[0164] Specifically, the pre-stored damage evolution model library is established by collecting a large amount of historical and effective data from slope monitoring. It includes three typical damage development stages: initial stability, damage development, and damage aggravation. Each stage corresponds to a reference feature vector. The reference trend of the initial stability model is quantified as 0 to represent stability, and the reference fluctuation amplitude is 0.1 to correspond to the baseline fluctuation of the stable stage. The reference trend of the damage development model is quantified as 0.5 to represent a slow rise, and the reference fluctuation amplitude is set to 0.3 to correspond to increased fluctuation during damage development. The reference trend of the damage aggravation model is quantified as 1.0 to represent a rapid rise, and the reference fluctuation amplitude is set to 0.6 to correspond to violent fluctuations during damage aggravation.

[0165] Specifically, the trend direction weight and volatility weight are set based on the importance of features. Normally, the trend direction is more critical than the volatility in determining the overall direction of damage evolution. Therefore, the trend direction weight is set to 0.6 and the volatility weight is set to 0.4, with the sum of the two being 1.

[0166] Specifically, the feature vector of the current analysis window is extracted, where the trend direction is the quantized value fitted within the window and the fluctuation range is the extracted standard deviation. The feature vector is matched with each reference feature vector in the pre-stored pattern library one by one. First, the reference trend direction and reference fluctuation range of each reference feature vector are obtained. The difference between the current trend direction and the reference trend direction are calculated, and the difference is squared and multiplied by the trend direction weight. The difference between the current fluctuation range and the reference fluctuation range is calculated, and the difference is squared and multiplied by the fluctuation range weight. The two results are summed and the square root of the sum is taken to obtain the weighted Euclidean distance.

[0167] In summary, this implementation method quantifies the similarity distance between the current feature and the reference feature. For example, the first distance is calculated by comparing the current feature vector's trend of 0.4 and fluctuation of 0.2 with the reference trend of 0 and fluctuation of 0.1 in the initial stable mode. The second distance is calculated by comparing the current feature vector's trend of 0.5 and fluctuation of 0.3 with the damage development mode. The third distance is calculated by comparing the current feature vector's trend of 1.0 and fluctuation of 0.6 with the damage aggravation mode. Comparing the three distances, the second distance is found to be the smallest. Therefore, the damage development stage corresponding to the damage development mode is determined as the most matching damage development stage.

[0168] In summary, this matching method can integrate the influence of two core features and avoid misjudgment based on a single feature. For example, looking only at the fluctuation amplitude may misjudge occasional fluctuations in the stable phase as damage. Combining the trend direction significantly improves the matching accuracy and can accurately determine the stage of damage development inside the rock mass.

[0169] In this embodiment of the invention, generating early warning information includes:

[0170] The corresponding warning level is determined by comparing the damage development stage with the preset warning level mapping relationship;

[0171] The warning level is combined with the current stage of damage development and key strongly correlated time periods to form the warning information.

[0172] Specifically, the preset warning levels are generally four levels: no warning, blue warning, orange warning, and red warning. The mapping relationship is set as follows: the initial stable stage corresponds to no warning, the damage development stage corresponds to blue warning, and the damage aggravation stage corresponds to orange warning. If the maximum value of the comprehensive state parameter in the damage aggravation stage exceeds the preset threshold, it will be upgraded to red warning. This threshold is derived based on the initial baseline value of 0.1 and is generally 5-10 times the baseline value. The threshold is set to 0.8.

[0173] Furthermore, after establishing the mapping table, the best matching damage development stage determined earlier is extracted. If there is a maximum value of the comprehensive state parameter, it is also checked whether it exceeds the threshold of 0.8. The damage development stage and related parameters are compared with the entries in the mapping table one by one to find the warning level corresponding to the completely matching entry, which is used as the current warning level.

[0174] Specifically, the current time is obtained directly through the system's built-in clock module to ensure the accuracy of time information. Key strongly correlated time period information includes the start and end times of the time period and the maximum value of the comprehensive state parameter within the time period. This information is extracted from the complete comprehensive state parameter sequence. The extraction process requires precise location of all marked strongly correlated time periods and selection of the time period most closely related to the current stage of damage development as the key strongly correlated time period.

[0175] Furthermore, the damage development stage at the current moment of the acquired warning level and the extracted key strongly correlated time period information are combined in a fixed format. For example, the fixed format is that the current moment is a certain time, the damage development stage inside the rock mass is determined to be a certain stage, the corresponding warning level is a certain level, the key strongly correlated time period is from a certain time period to a certain time period, and the maximum value of the comprehensive state parameter within that time period is a certain value. This combination of actions can completely integrate the core monitoring information, and the resulting warning information contains comprehensive key data.

[0176] In general, staff can quickly pinpoint the root cause of a problem by using the damage stage and strongly correlated time periods in the early warning information, and take targeted reinforcement or monitoring enhancement measures, which can significantly improve the accuracy and efficiency of the early warning response.

[0177] In the several embodiments provided by this invention, it should be understood that the disclosed method can be implemented in other ways.

[0178] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0179] The embodiments of this application can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence is the theory, method, and technology that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0180] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. An active waveguide acoustic emission monitoring method for slope stability monitoring, characterized in that, The method includes: S1. Drill holes in the slope monitoring area and insert waveguide rods with active excitation elements, then inject grout to solidify and form a rigid integrated waveguide sensing structure that is acoustically coupled with the rock mass. S2. By using a rigid integrated waveguide sensing structure, the passive signal of rock micro-fracture and the active reference signal triggered by the active excitation element are simultaneously acquired to obtain a mixed signal. S3. Using the inherent stable waveform of the active reference signal in the rigid integrated waveguide sensing structure as a reference, the signal substrate is separated from the synchronously acquired signal, and the passive signal is extracted after the signal substrate is removed. S4. Extract the break band parameter modulated by the conduction characteristics of the rigid integrated waveguide sensing structure from the passive signal; at the same time, extract the propagation timing parameter and amplitude parameter characterizing the physical state change of the rigid integrated waveguide sensing structure itself from the active reference signal. The two together constitute the active propagation state parameter. S5. Jointly analyze the rupture frequency band parameters and the active propagation state parameters, establish their change correlation, and generate comprehensive state parameters; S6. Based on the temporal evolution pattern of the comprehensive state parameters, determine the stress state perceived by the rigid coupling body and the internal damage development of the rock mass, and generate early warning information. The step of separating the signal substrate from the synchronously acquired signal, based on the inherent stable waveform of the active reference signal in the rigid integrated waveguide sensing structure, includes: Acquire the waveform of the active reference signal within a preset stabilization period and establish a reference waveform template; Perform waveform matching operation between the mixed signal and the reference waveform template to locate the part of the mixed signal whose shape similarity to the reference waveform template meets the preset conditions; The parts whose morphological similarity meets the preset conditions are removed from the mixed signal to obtain the residual signal after substrate removal; The joint analysis of the rupture frequency band parameters and the active propagation state parameters to establish their change correlation includes: Synchronize the rupture frequency band parameter sequence and the active propagation state parameter sequence on the time axis; Identify and compare the points of change in the rupture frequency band parameters with the points of change in the active propagation state parameters; The time intervals in which the rupture frequency band parameter and the active propagation state parameter both change significantly are marked as strong correlation periods; The generated comprehensive state parameters include: During periods of strong correlation, preset weights are assigned to the intensity of changes in the rupture frequency band parameter and the intensity of changes in the active propagation state parameter, respectively. The two change intensity values, after being assigned preset weights, are superimposed to obtain the comprehensive state parameter value for that period. The comprehensive state parameter values ​​for all time periods are concatenated in chronological order to form a complete comprehensive state parameter sequence.

2. The active waveguide acoustic emission monitoring method for slope stability monitoring as described in claim 1, characterized in that, The process of extracting the passive signal after removing the signal substrate includes: Time-domain energy peak detection is performed on the residual signal to identify all transient pulse events that exceed the energy threshold; The time-frequency distribution characteristics of transient pulse events are compared with a pre-defined rupture signal feature library, and pulse events with matching characteristics are selected as passive signals.

3. The active waveguide acoustic emission monitoring method for slope stability monitoring as described in claim 1, characterized in that, The extraction of the broken frequency band parameters modulated by the conduction characteristics of the rigid integrated waveguide sensing structure from the passive signal includes: Perform spectral transformation on each event in the passive signal to obtain the spectrum of each event; From the spectrum of each event, locate the continuous frequency band with the highest energy concentration, and determine the center position and bandwidth parameters of the frequency band; Arrange the frequency band center position and width parameters corresponding to each event in chronological order to form a sequence of frequency band break parameters.

4. The active waveguide acoustic emission monitoring method for slope stability monitoring as described in claim 1, characterized in that, The extraction of propagation timing parameters and amplitude parameters characterizing the physical state changes of the rigid integrated waveguide sensing structure itself from the active reference signal includes: Identify the start time of each active reference signal event and determine its propagation time in the rigid integrated waveguide sensing structure; Measure the peak amplitude of each active reference signal event; The propagation timing variation parameter is determined by the difference in propagation duration between adjacent events; the amplitude variation parameter is determined by the ratio of the peak amplitudes of adjacent events.

5. The active waveguide acoustic emission monitoring method for slope stability monitoring as described in claim 4, characterized in that, The two together constitute the active propagation state parameters, including: Align the propagation timing variation parameter with the amplitude variation parameter using the same time coordinate; The propagation time sequence change value and amplitude change value corresponding to each moment are combined into a state vector; Arrange the state vectors at all times in chronological order to form an active propagation state parameter sequence.

6. The active waveguide acoustic emission monitoring method for slope stability monitoring as described in claim 1, characterized in that, The determination of internal damage development in the rock mass based on the temporal evolution pattern of comprehensive state parameters includes: The integrated state parameter sequence is divided into multiple consecutive analysis windows; Within each analysis window, extract the trend and fluctuation characteristics of the comprehensive state parameter values; The trend and fluctuation characteristics are matched with a pre-stored damage evolution pattern library to determine the most matching damage development stage.

7. The active waveguide acoustic emission monitoring method for slope stability monitoring as described in claim 6, characterized in that, The generation of early warning information includes: The corresponding warning level is determined by comparing the damage development stage with the preset warning level. The warning level is combined with the current stage of damage development and key strongly correlated time periods to form the warning information.

Images