Method for checking dangerous rock mass stability evaluation result based on historical data comparison

Abstract

The present application relates to the technical field of geological engineering, in particular to a dangerous rock mass stability evaluation result verification method based on historical data comparison, comprising the following steps: obtaining dangerous rock mass frequency and mechanical monitoring data, dividing stable and critical working conditions, constructing stable and critical state offset difference surface and energy barycenter coupling difference surface, analyzing seismic motion trigger frequency and instability frequency to form load trigger comparison information, calculating crack propagation rate and displacement variation and quantifying stage state, and outputting stage joint verification data, in the present application, through frequency slope identification, mechanical parameter offset depiction, frequency energy and barycenter direction coupling, load trigger and instability proportion offset, crack and displacement joint quantification, continuous determination of stage state is realized, critical signal identification sensitivity is improved, multi-source evidence consistency is enhanced, objectivity and reproducibility of result verification are optimized, experience dependence is reduced, and stability evaluation and observation response are closely corresponded.

Description

Technical Field

[0001] This invention relates to the field of geological engineering technology, and in particular to a method for verifying the stability evaluation results of unstable rock masses based on historical data comparison. Background Technology

[0002] The field of geological engineering technology involves the study of the formation, evolution, and engineering stability of geological bodies. It primarily includes core aspects such as rock mass structural characteristic analysis, geostress measurement, research on the deformation and failure mechanisms of rock and soil masses, slope stability evaluation, and geological disaster prevention. This field's technical system integrates theories from multiple disciplines, including geology, rock mechanics, structural dynamics, and engineering monitoring. Through field surveys, model calculations, and monitoring data analysis, it comprehensively assesses the mechanical response and stability of rock masses under natural or human influences, providing a scientific basis for engineering design, disaster early warning, and risk management. Among these methods, the traditional verification method for the stability evaluation results of dangerous rock masses refers to... The safety factor, dynamic safety factor, and instability probability calculated in rock mass stability analysis are verified for accuracy through manual or empirical means. Traditionally, this verification relies on historical monitoring records, field investigations, and case comparisons. Specifically, this includes collecting deformation displacement, crack propagation, and collapse records of known unstable rock masses, comparing the consistency between the predicted values ​​of the calculation model and the measured data, reviewing rock mass parameters in typical historical instability events to determine the reliability of the model analysis, and combining seismic activity statistics and surface deformation trends for qualitative verification to reconfirm the model calculation results and ensure the scientific rationality of the unstable rock mass stability evaluation results.

[0003] Traditional verification of rock mass stability assessment results relies on manual experience to compare monitoring records and typical cases, and on consistency judgment based on event-driven deformation, crack propagation, or historical collapse records. Monitoring data is mostly used on long-term trend scales, lacking processing of temporal continuity and fine-grained dynamic responses. This weakens the correlation between slope changes, energy migration, and mechanical parameter shifts, making it difficult to identify hidden critical state signals from the monitoring sequence. The connection between triggering loads and instability behavior relies on macroscopic statistical data, which cannot reflect the differences between different stages. As a result, the verification process lacks quantitative basis for stage characteristics, affecting the credibility of stability assessment results. Summary of the Invention

[0004] To address the technical problems existing in the prior art, embodiments of the present invention provide a method for verifying the stability evaluation results of unstable rock masses based on historical data comparison, comprising the following steps:

[0005] To achieve the above objectives, the present invention adopts the following technical solution: a method for verifying the stability evaluation results of unstable rock masses based on historical data comparison, comprising the following steps:

[0006] S1: Obtain the frequency monitoring sequence of the unstable rock mass, analyze the continuous changes over time, calculate the slope of the frequency change and obtain the slope difference, identify the period of significant dynamic change, divide the stable segment and the attenuation segment, divide the monitoring sequence into multiple dynamic behavior categories, and generate working condition slope change amplitude groups.

[0007] S2: Based on the slope change range group of the working conditions, analyze the rock mass quality, structural surface dip angle, cohesion and internal friction angle sequence of the corresponding period, perform differential operation on the mechanical parameters to form an offset sequence, identify the offset inflection point and boundary period, construct the difference data between steady state and critical state, and generate the offset difference surface between steady state and critical state.

[0008] S3: Based on the steady-state and critical-state offset difference surface, calculate the acceleration spectrum energy proportion distribution, determine the consistency between the spectrum centroid direction and the energy migration direction, summarize the directional differences within the steady-state and critical-state segments, and generate the energy centroid coupling difference surface;

[0009] S4: Based on the energy centroid coupling difference surface, analyze the peak ground motion sequence, count the triggering frequency in segments according to the triggering benchmark, calculate the occurrence frequency in combination with the monitoring instability sequence, compare the proportional offset of the two types of frequencies and integrate them across periods to generate load triggering comparison information.

[0010] S5: Based on the load trigger comparison information, calculate the deformation increment and expansion rate of the same unstable rock mass crack width sequence, filter the time point corresponding to the rate inflection point, locate the corresponding time of the displacement sequence and take the displacement change and cumulative change, combine normalization processing, quantify the stage state strength of the unstable rock mass, and generate stage joint verification data.

[0011] As a further embodiment of the present invention, the slope variation range group includes a stable slope range, a decay slope range, and a transition slope range; the steady-state and critical-state offset difference surface includes a steady-state parameter offset field, a critical-state parameter offset field, and a steady-state-critical-state transition offset gradient field; the energy centroid coupling difference surface includes an energy proportion variation field, a spectral centroid drift field, and a distribution of energy and centroid direction coupling coefficients; the load triggering comparison information includes a seismic ground motion triggering frequency distribution, an instability event response frequency distribution, and a triggering probability offset distribution field; and the stage joint verification data includes a crack propagation rate sequence, a displacement variation sequence, and a stage state intensity quantification sequence.

[0012] As a further aspect of the present invention, the step of obtaining the slope change amplitude group under the working conditions specifically includes:

[0013] S101: Obtain the frequency monitoring sequence of the unstable rock mass, detect the frequency corresponding to each time period on the time axis and the time difference between adjacent time periods, calculate the frequency change slope of adjacent time periods, screen time slices with significant slope changes, and establish a dynamic change time slice group.

[0014] S102: Based on the dynamic change time slice group, compare the slope sequence before and after the target time slice, calculate the slope change amplitude of each time period, determine the numerical range of the slope change amplitude, divide the stable segment and the decay segment, and generate a dynamic behavior interval group.

[0015] S103: Based on the dynamic behavior interval group, collect the slope change amplitude of multiple time periods, classify and organize the dynamic behavior categories, and generate the working condition slope change amplitude group.

[0016] As a further aspect of the present invention, the process of determining the numerical range of the slope change amplitude and dividing the stable segment and the attenuation segment is specifically as follows:

[0017] Using the equal time intervals between adjacent dynamic change time slices and without dynamic change time slices as reference intervals, the median and dispersion of the slope change amplitude sequence within the reference intervals are calculated respectively, and the smaller of the medians of the two reference intervals is used as the stability benchmark of the slope change amplitude.

[0018] The slope change amplitude sequence within the dynamic change time slice is used as the upper quantile, and the slope change amplitude stability benchmark is combined to set the slope change amplitude attenuation judgment threshold.

[0019] The slope change amplitude sequence is compared with the slope change amplitude stability benchmark and the slope change amplitude attenuation judgment threshold in chronological order. When the slope change amplitude is between the slope change amplitude stability benchmark and the slope change amplitude attenuation judgment threshold, it is divided into a stable segment. When the slope change amplitude exceeds the slope change amplitude attenuation judgment threshold, it is divided into an attenuation segment.

[0020] As a further aspect of the present invention, the step of obtaining the steady-state and critical-state shift difference surface specifically comprises:

[0021] S201: Based on the slope change range group of the working conditions, obtain the rock mass quality, structural surface dip angle, cohesion and internal friction angle sequence for the corresponding period, perform differential operation on each mechanical parameter at the same time point, and establish a multi-parameter offset sequence;

[0022] S202: Based on the multi-parameter offset sequence, compare the composite change amplitude of each parameter, identify the offset inflection point and the corresponding boundary time period, compare the composite change amplitude before and after the time period, and obtain the offset boundary interval data.

[0023] S203: Based on the offset boundary interval data, calculate the concentration and variation amplitude of the offset sequence before and after the boundary, construct the difference data between the steady state and the critical state, and generate the offset difference surface between the steady state and the critical state.

[0024] As a further aspect of the present invention, the process of identifying the offset inflection point and the corresponding boundary time period is specifically as follows:

[0025] Using the time periods at the beginning and end of the time axis in the multi-parameter offset sequence that do not show abrupt changes in the composite change amplitude as reference segments, the median and dispersion of the composite change amplitude at each time point within the reference segments are calculated, and the reference range of the composite change amplitude is set based on the median and dispersion.

[0026] The median of the composite change range of the multi-parameter offset sequence over the entire time range is used as the candidate baseline for the offset inflection point, and the candidate baseline for the offset inflection point is jointly determined with the reference range of the composite change range to determine the threshold for the offset inflection point.

[0027] The composite change amplitude and the offset inflection point determination threshold of each time point are compared sequentially according to the multi-parameter offset sequence. When the composite change amplitude continuously crosses the offset inflection point determination threshold and maintains a preset duration on both sides of the offset inflection point determination threshold, the time point corresponding to the position where the composite change amplitude crosses the offset inflection point determination threshold is determined as the offset inflection point, and the time segments before and after the offset inflection point that meet the preset duration conditions are determined as the corresponding boundary time periods.

[0028] The preset duration is the time length corresponding to three consecutive monitoring time points.

[0029] As a further aspect of the present invention, the step of obtaining the energy centroid coupling difference surface specifically includes:

[0030] S301: Based on the difference surface between steady state and critical state offset, obtain the acceleration monitoring data for the corresponding time period, perform spectral transformation on the acceleration sequence for each time period, divide the spectrum into multiple frequency bands, and perform energy accumulation calculation on each frequency band to establish the frequency band energy proportion distribution;

[0031] S302: Based on the energy proportion distribution of the frequency band, identify the direction of energy proportion change in adjacent time periods, calculate the position change of the centroid of the spectrum in each time period, determine the consistency between the energy direction and the centroid direction in the time series, and establish a direction consistency sequence.

[0032] S303: Call the aforementioned direction consistency sequence, summarize the direction differences in the steady-state segment and the critical state segment respectively, organize the difference index of the coupling direction, and generate the energy centroid coupling difference surface.

[0033] As a further aspect of the present invention, the step of obtaining the load trigger comparison information specifically includes:

[0034] S401: Based on the energy centroid coupling difference surface, analyze the peak ground motion sequence of the corresponding period, judge according to the triggering benchmark, count the occurrence frequency of ground motion triggering events in each segment, and generate earthquake triggering frequency data.

[0035] S402: Based on the earthquake triggering frequency data, obtain the instability event sequence in the monitoring records of unstable rock masses during the same period, count the occurrence frequency of instability events, perform proportional offset calculation on the earthquake triggering frequency and the instability event frequency, and obtain probability offset amplitude data;

[0036] S403: Based on the probability offset amplitude data, evaluate the overall offset intensity based on the probability offset amplitude of multiple periods, organize the relationship between corresponding load triggering and probability offset, and generate load triggering comparison information.

[0037] As a further aspect of the present invention, the step of obtaining the stage joint verification data specifically includes:

[0038] S501: Based on the load triggering comparison information, according to the time series of the same unstable rock mass crack width, perform deformation increment calculation on the width of adjacent time periods to obtain the crack propagation rate sequence, filter the time points corresponding to the rate inflection points, and generate crack rate feature point data.

[0039] S502: Call the fracture rate feature point data, extract the corresponding time points in the displacement sequence, and perform synchronous extraction on the displacement change and cumulative change for each time period to obtain a fracture and displacement variation data set.

[0040] S503: Based on the fracture and displacement variation data set, the fracture propagation rate, displacement change, cumulative change and load trigger comparison information are normalized to quantify the stage state strength of the unstable rock mass and generate stage joint verification data.

[0041] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0042] This invention takes the continuous changes in the monitoring sequence as the starting point, and constructs a multi-layered verification chain covering dynamic response, mechanical offset, frequency domain coupling, and load correlation by calculating the slope of frequency change, mechanical parameter offset, energy proportion migration direction, and linkage quantification of cracks and displacement. It uses slope difference to locate periods of significant dynamic change, multi-parameter difference to identify critical boundaries, energy and center of gravity direction comparison to characterize the difference between stability and instability, and proportional offset of triggering behavior and instability behavior to form load correlation basis. Combined with deformation increment and expansion rate, it completes the stage state quantification, realizing continuous, fine-grained and physical response-based verification of stability evaluation results, solving the problems of trend-based, experience-based, and lack of stage support in traditional verification. Attached Figure Description

[0043] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0044] Figure 1 This is a schematic diagram of the steps of the present invention;

[0045] Figure 2 This is a detailed schematic diagram of S1 of the present invention;

[0046] Figure 3 This is a detailed schematic diagram of S2 of the present invention;

[0047] Figure 4 This is a detailed schematic diagram of S3 of the present invention;

[0048] Figure 5 This is a detailed schematic diagram of S4 of the present invention;

[0049] Figure 6 This is a detailed schematic diagram of S5 of the present invention. Detailed Implementation

[0050] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0051] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0052] In the embodiments of this invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning. Similarly, the terms "of," "corresponding (relevant)," and "corresponding" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning.

[0053] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0054] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0055] Please see Figure 1 This invention provides a method for verifying the stability evaluation results of unstable rock masses based on historical data comparison, including the following steps:

[0056] S1: Obtain the frequency monitoring sequence of the unstable rock mass, analyze the continuous changes over time, calculate the slope of the frequency change and obtain the slope difference, identify the period of significant dynamic change, divide the stable segment and the attenuation segment, divide the monitoring sequence into multiple dynamic behavior categories, and generate working condition slope change amplitude groups.

[0057] S2: Based on the slope variation range group of working conditions, analyze the rock mass quality, structural surface dip angle, cohesion and internal friction angle sequence of the corresponding period, perform differential calculation on the mechanical parameters to form the migration sequence, identify the migration inflection point and boundary period, construct the difference data between steady state and critical state, and generate the migration difference surface between steady state and critical state.

[0058] S3: Based on the offset difference surface between steady state and critical state, calculate the energy proportion distribution of the acceleration spectrum, determine the consistency between the direction of the spectrum centroid and the direction of energy migration, summarize the directional differences in the steady state and critical state segments, and generate the energy centroid coupling difference surface;

[0059] S4: Based on the energy center of gravity coupling difference surface, analyze the peak ground motion sequence, count the triggering frequency in segments according to the triggering benchmark, calculate the occurrence frequency in combination with the monitoring instability sequence, compare the proportional offset of the two types of frequencies and integrate them across periods to generate load triggering comparison information.

[0060] S5: Based on load-triggered comparison information, calculate the deformation increment and propagation rate of the same unstable rock mass crack width sequence, filter the time points corresponding to the rate inflection points, locate the corresponding time of the displacement sequence and take the displacement change and cumulative change, combine normalization processing, quantify the stage state strength of the unstable rock mass, and generate stage joint verification data.

[0061] The slope variation range group includes the slope range of the stable segment, the slope range of the attenuation segment, and the slope range of the transition segment. The difference surface between steady state and critical state offset includes the steady state parameter offset field, the critical state parameter offset field, and the steady-to-critical transition offset gradient field. The difference surface between energy and centroid coupling includes the energy proportion variation field, the spectral centroid drift field, and the distribution of the coupling coefficient between energy and centroid direction. The load triggering comparison information includes the distribution of ground motion triggering frequency, the distribution of instability event response frequency, and the triggering probability offset distribution field. The stage joint verification data includes the crack propagation rate sequence, the displacement variation sequence, and the stage state intensity quantification sequence.

[0062] Please see Figure 2 The specific steps for obtaining the slope change amplitude group under operating conditions are as follows:

[0063] S101: Obtain the frequency monitoring sequence of the unstable rock mass, detect the frequency corresponding to each time period on the time axis and the time difference between adjacent time periods, calculate the frequency change slope of adjacent time periods, screen time slices with significant slope changes, and establish a dynamic change time slice group.

[0064] For a specific unstable rock mass GTC-01, distributed fiber optic sensors installed inside the rock mass collect its natural frequency at a 1-hour cycle. First, a frequency monitoring sequence of 120 time periods is acquired from 00:00 on January 1, 2024 to 23:00 on January 5, 2024. Each element in the sequence contains the monitoring time point and the corresponding natural frequency value. For example, the first three data points of the sequence are (2024-01-01-00:00, 4.502Hz), (2024-01-01-01:00, 4.505Hz), and (2024-01-01-02:00, 4.506Hz). For each time period in the sequence, its corresponding frequency value and the time difference with the adjacent previous time period are detected. In this embodiment, the time difference is a fixed monitoring period, i.e., Δt = 1 hour. Subsequently, the slope of the frequency change between each pair of adjacent time periods is calculated. This is done by subtracting the frequency value of the previous time period from the frequency value of the later time period, and then dividing by the time difference between the two time periods. Taking the second time period as an example, its frequency change slope is... Hz / hour. Following this method, calculations were performed on all 119 calculable intervals to obtain a time series containing 119 slope values. The next step is to screen time slices with significant slope changes. To determine the quantitative standard for "significance," a threshold must first be set. This threshold is based on a statistical analysis of 8759 frequency change slope values ​​calculated from 8760 monitoring data points for the GTC-01 rock mass over the past year (the entire year of 2023). The arithmetic mean and standard deviation of these 8759 slope values ​​were calculated; the mean was 0.0001 Hz / hour, and the standard deviation was 0.0025 Hz / hour. The significance threshold was determined to be the mean plus twice the standard deviation, i.e. Hz / hour. This threshold was selected based on retrospective analysis of historical data. It was found that in three past minor rockfall events, the absolute value of the frequency change slope consistently exceeded this threshold within the 24 hours preceding each event, achieving a correlation accuracy of 87.5%. Subsequently, the absolute values ​​of 119 slope values ​​calculated during the current monitoring period (January 1st to January 5th) were compared one by one with the threshold of 0.0051 Hz / hour. The comparison revealed that in five time periods—from the 70th period (corresponding to 21:00 to 22:00 on January 3rd, 2024) to the 74th period (corresponding to 01:00 to 02:00 on January 4th, 2024)—the absolute values ​​of the slopes were 0.0058, 0.0065, 0.0072, 0.0068, and 0.0055 Hz / hour, respectively, all exceeding 0.0051 Hz / hour. Therefore, the five-hour time interval from 21:00 on January 3, 2024 to 02:00 on April 4, 2024 is selected and determined as a dynamic change time slice. This process is repeated to complete the selection of the entire monitoring sequence, and finally a dynamic change time slice group is established, which in this example contains one element: [2024-01-03-21:00, 2024-01-04-02:00].

[0065] S102: Based on the dynamic change time slice group, compare the slope sequence before and after the target time slice, calculate the slope change amplitude of each time period, determine the numerical range of the slope change amplitude, divide the stable segment and the decay segment, and generate the dynamic behavior interval group.

[0066] Based on the dynamic change time slice group, a unique element, [2024-01-03 21:00, 2024-01-04 02:00], is selected as the target time slice. To distinguish between the stable and decaying segments, reference segments need to be defined and corresponding discrimination criteria and thresholds set. First, the equal-length segments adjacent to the dynamic change time slice (5 hours in duration) but excluding it are used as reference segments. Specifically: Reference segment one, from 2024-01-03 16:00 to 2024-01-03 21:00; Reference segment two, from 2024-01-04 02:00 to 2024-01-04 07:00. The frequency change slope sequence (each segment contains 5 slope values) within these two reference segments is calculated, i.e., the absolute value of the slope, resulting in two amplitude sequences. The slope variation amplitude sequence for reference segment one is [0.0012, 0.0008, 0.0015, 0.0011, 0.0013] Hz / hour, and the slope variation amplitude sequence for reference segment two is [0.0018, 0.0015, 0.0021, 0.0016, 0.0019] Hz / hour. The median and dispersion (interquartile range, IQR) are calculated for both sequences. The median for reference segment one is 0.0012 Hz / hour, and the dispersion is 0.0004 Hz / hour. The median for reference segment two is 0.0018 Hz / hour, and the dispersion is 0.0005 Hz / hour. The smaller of the two medians, 0.0012 Hz / hour, is taken as the slope variation amplitude stability benchmark. Next, a threshold for determining slope variation amplitude attenuation is set. First, the slope variation amplitude sequence within the dynamic variation time slice (2024-01-03-21:00 to 2024-01-04-02:00) is obtained, with values ​​of [0.0058, 0.0065, 0.0072, 0.0068, 0.0055] Hz / hour. The upper quartile (Q3) of this sequence is calculated, yielding 0.0070 Hz / hour. Combining this with the aforementioned slope variation amplitude stability benchmark of 0.0012 Hz / hour, the attenuation judgment threshold is set as follows: The coefficient C was determined through retrospective analysis of 10 sets of known unstable rock mass data from historical datasets. Tests were conducted within the C value range [1.0, 2.0] with a step size of 0.1. When C=1.5, the accuracy rate for identifying the attenuation segment was the highest, reaching 91%. Therefore, in this embodiment, C is set to 1.5. Thus, the attenuation judgment threshold is calculated as follows: Hz / hour. Finally, for the entire monitoring period (16:00 on January 3, 2024 to 07:00 on January 4, 2024), including the reference segment and the dynamic change time slice, the slope change amplitude sequence was compared interval by interval in chronological order. When the slope change amplitude value of a time segment was within the interval (0.0012, 0.0099] Hz / hour, that time segment was classified as a stable segment; when it exceeded 0.0099 Hz / hour, it was classified as a decaying segment. In this example, the slope change amplitude values ​​of all 15 time segments did not exceed 0.0099 Hz / hour, but the values ​​of 10 time segments were greater than 0.0012 Hz / hour, so these time segments were all classified as stable segments, and there were no decaying segments. This generated a dynamic behavior interval group, that is, a time series marked as a "stable segment".

[0067] S103: Based on the dynamic behavior interval group, collect the slope change amplitude of multiple time periods, classify and organize the dynamic behavior categories, and generate the working condition slope change amplitude group.

[0068] Based on the dynamic behavior interval groups, namely a series of time segments marked as "stable segments" or "decaying segments" and their corresponding slope change amplitude values, the data is collected and classified. This step aims to abstract the scattered, time-sequential dynamic behaviors into statistically characteristic operating condition categories. The operation process is as follows: the slope change amplitude values ​​of all time slices classified as "stable segments" within the monitoring period (in this example, January 1 to January 5, 2024) are collected into one set, denoted as set A. Similarly, the slope change amplitude values ​​of all time slices classified as "decaying segments" are collected into another set, denoted as set B. In the example of S102, only "stable segments" were identified, and no "decaying segments" appeared, so set B is empty. To make the embodiment complete, another segment of historical monitoring data is introduced, in which a decaying segment is identified, with a slope change amplitude value of [0.0102, 0.0115, 0.0108] Hz / hour. These three values ​​constitute set B. Set A contains the amplitude values ​​of the 10 stable segments identified in S102, as well as the amplitude values ​​of 15 other stable segments from historical data, totaling 25 values. The next step is to classify and organize this aggregated data into dynamic behavior categories. Category 1 is defined as "steady-state operating condition," and its data comes from set A. Category 2 is defined as "decaying operating condition," and its data comes from set B. Finally, a set of slope change amplitudes is generated for each type of dynamic behavior, which essentially establishes a feature description for each type of operating condition. This description is achieved by calculating the statistical parameters of all slope change amplitude values ​​under that category. For "steady-state operating condition" (Category 1), the mean, median, standard deviation, and coefficient of variation of the 25 amplitude values ​​in set A are calculated. The calculated mean is 0.0045 Hz / hour, the median is 0.0041 Hz / hour, the standard deviation is 0.0018 Hz / hour, and the coefficient of variation is 0.40. For the "attenuation condition" (Category 2), calculate the mean, median, standard deviation, and coefficient of variation for the three amplitude values ​​in set B. The calculated mean is 0.0108 Hz / hour, the median is 0.0108 Hz / hour, the standard deviation is 0.00066 Hz / hour, and the coefficient of variation is 0.061. The resulting set of amplitude values ​​representing the slope changes of the operating conditions represents the two sets of statistical characteristic parameters describing the steady-state and attenuation conditions respectively: {Steady-state condition: {Mean: 0.0045, Median: 0.0041, Standard deviation: 0.0018, Coefficient of variation: 0.40}, Attenuation condition: {Mean: 0.0108, Median: 0.0108, Standard deviation: 0.00066, Coefficient of variation: 0.061}}.

[0069] Please see Figure 3 The specific steps for obtaining the difference surface between steady-state and critical-state shifts are as follows:

[0070] S201: Based on the slope variation range group of the working conditions, obtain the sequence of rock mass quality, structural surface dip angle, cohesion and internal friction angle for the corresponding period, perform differential operation on each mechanical parameter at the same time point, and establish a multi-parameter offset sequence;

[0071] Based on the different working condition time periods defined by the slope variation range group, namely the time intervals corresponding to "steady-state working condition" and "attenuation working condition", the multi-source mechanical parameter monitoring sequence of the unstable rock mass GTC-01 during these same periods was obtained. These parameters were obtained through sensors installed inside and on the surface of the rock mass and through regular field tests. The specific sequences obtained include: 1) Rock mass quality, here using the Geological Strength Index (GSI), automatically generated daily by the rock mass structure detector, with a value range of 0-100; 2) Structural plane dip angle, measured hourly by an automated inclinometer, in degrees (°); 3) Structural plane cohesion, obtained through periodic (daily) direct shear tests on samples collected from the rock mass surface, in kilopascals (kPa); 4) Structural plane internal friction angle, obtained through the same set of tests as cohesion, in degrees (°). For example, on January 3, 2024, the following partial data sequences were obtained: GSI sequence [55.2, 55.1, 55.1, 54.9, ...]; tilt angle sequence [30.15, 30.16, 30.18, 30.20, ...] degrees; cohesion sequence [850.5, 850.2, 849.8, 849.1, ...] kPa; internal friction angle sequence [28.5, 28.5, 28.4, 28.4, ...] degrees. After obtaining these sequences, for each mechanical parameter, a difference operation was performed at the same time point, that is, the parameter value at the current time was subtracted from the parameter value at the previous time. Taking 02:00 on January 3, 2024 as an example, assuming that the GSI was 55.1 at 01:00 and 55.1 at 02:00, then the GSI offset value at that time is... Assuming the tilt angle is 30.16° at 01:00 and 30.18° at 02:00, then the tilt angle offset is... Assuming the cohesion is 850.2 kPa at 01:00 and 849.8 kPa at 02:00, then the cohesion offset value is... kPa. Assuming the internal friction angle is 28.5° at 01:00 and 28.4° at 02:00, then the internal friction angle offset value is... This difference operation is repeated for all time points within the entire monitoring period, ultimately establishing a time series for each of the four parameters: GSI, structural surface tilt angle, cohesion, and internal friction angle. These four series together constitute a multi-parameter migration sequence.

[0072] S202: Based on the multi-parameter offset sequence, compare the composite change amplitude of each parameter, identify the offset inflection point and the corresponding boundary time period, compare the composite change amplitude before and after the time period, and obtain the offset boundary interval data.

[0073] Based on the multi-parameter offset sequences, namely the GSI offset sequence, tilt angle offset sequence, cohesion offset sequence, and internal friction angle offset sequence, a comparison of the synthesized change amplitude is performed to identify the offset inflection point. First, due to the significant differences in the dimensions and numerical ranges of the offset values ​​of each parameter, normalization is required. Here, the maximum-minimum normalization method is used. For each offset sequence, its value range is mapped to an interval based on its maximum and minimum values ​​in the long-term monitoring history. For example, if the historical maximum absolute value of cohesion offset is 5.0 kPa, then the normalized absolute value of cohesion offset is... After normalizing all four offset sequences, the composite change amplitude at each time point is calculated. This amplitude is defined as the weighted sum of the absolute values ​​of the four normalized offset values: Among them, the weighting coefficient The determination was made through the Analytic Hierarchy Process (AHP), in which five experts in the field of geological engineering conducted pairwise comparisons and scoring of the importance of each parameter to the stability of the unstable rock mass. The result is as follows: , , , For example, at time t, the absolute values ​​of the four normalized offsets are 0.01, 0.05, 0.08, and 0.02, respectively. Then the composite variation amplitude at that time is... Next, the threshold for determining the offset inflection point is set. First, a two-week period (one week at the beginning and one week at the end of the timeline) without any sudden changes in the composite change amplitude (defined as an amplitude value exceeding the historical 95th percentile) is selected as the reference segment. The median (0.025) and interquartile range (IQR=0.018) of all composite change amplitude values ​​within these two segments are calculated, setting the baseline range to [0.007, 0.043]. Then, the median (0.048) of all composite change amplitudes throughout the entire monitoring period is calculated as the candidate baseline for the offset inflection point. Finally, the offset inflection point determination threshold is set as the combination of the candidate baseline and the upper limit of the baseline range, i.e. The coefficient of 0.5 is obtained based on historical data backtesting. At this value, the F1 score for identifying state transition points is the highest. Finally, the composite change amplitude at each time point is compared with the threshold of 0.057 in chronological order. When it is found that the composite change amplitude changes from below 0.057 to above 0.057 for three consecutive time points (preset duration), and maintains the state for at least three time points on both sides of the threshold, the first time point that crosses the threshold is determined as the offset inflection point. Assuming that the offset inflection point is determined at 15:00 on January 20, 2024, the three-hour intervals before and after it (12:00-15:00 and 15:00-18:00) are the corresponding boundary time periods. Thus, the offset boundary interval data is obtained.

[0074] S203: Based on the offset boundary interval data, calculate the concentration and variation amplitude of the offset sequence before and after the boundary, construct the difference data between the steady state and the critical state, and generate the offset difference surface between the steady state and the critical state;

[0075] Based on the inflection point (2024-01-20 15:00) and the corresponding boundary time periods (first segment: 12:00-15:00; second segment: 15:00-18:00), a differential analysis was performed on the mechanical parameter migration sequences of these two segments. The purpose of this step is to quantify the changes in the internal mechanical parameter response patterns of the rock mass as it transitions from a "steady state" to a "critical state." First, the migration sequences of four mechanical parameters were extracted for the first segment (considered the steady-state region) and the second segment (considered the critical state region). For each parameter migration sequence within each segment, two core indicators were calculated: concentration and variation amplitude. Concentration is defined here as the reciprocal of the sequence standard deviation, i.e. A larger value indicates that the data points are more concentrated around the mean. The amplitude of change is defined as the mean of the absolute values ​​in the sequence, i.e. This value reflects the average drastic degree of parameter change. Taking cohesion (c) as an example, in the front section of the boundary (steady-state region), its offset sequence is [-0.1, -0.2, 0.1] kPa, with a mean of -0.067 kPa, a standard deviation of 0.153 kPa, and an absolute mean of 0.133 kPa. Therefore, the cohesion concentration in the steady-state region is... The variation range is 0.133 kPa. In the later part of the boundary (critical state region), its offset sequence is [-0.8, -1.2, -1.0] kPa, with a mean of -1.0 kPa, a standard deviation of 0.2 kPa, and an absolute mean of 1.0 kPa. Therefore, the cohesion concentration in the critical state region is... The variation range is 1.0 kPa. This calculation process is applied to all four mechanical parameters (GSI, tilt angle, cohesion, and internal friction angle) to obtain their concentration and variation range in the steady-state and critical states, respectively.

[0076] Table 1. Data on differences in steady-state mechanical parameters

[0077]

[0078] As shown in Table 1, by comparing these indicators, the difference data between steady state and critical state were constructed. These data were integrated to generate a structured dataset, namely the steady-state and critical state offset difference surface. This difference surface clearly reveals that from steady state to critical state, the offset of all mechanical parameters exhibits a common trend of "decreasing concentration" (i.e., data is more discrete and fluctuations are more irregular) and "increasing amplitude of change" (i.e., the parameter deterioration rate is faster).

[0079] Please see Figure 4 The specific steps for obtaining the energy centroid coupling difference surface are as follows:

[0080] S301: Based on the difference surface between steady-state and critical-state offsets, obtain acceleration monitoring data for the corresponding time period, perform spectral transformation on the acceleration sequence for each time period, divide the spectrum into multiple frequency bands, perform energy accumulation calculation on each frequency band, and establish the frequency band energy proportion distribution.

[0081] Based on the steady-state and critical-state migration difference surface, determining the steady-state region (2024-01-20-12:00-15:00) and critical-state region (2024-01-20-15:00-18:00), acceleration monitoring data collected by microseismic sensors during these two time periods were obtained from the monitoring system of the GTC-01 unstable rock mass. Acceleration data was sampled at 0.01-second intervals, with units of... For each time-series acceleration data point, a spectral transformation is performed. Specifically, a Fast Fourier Transform (FFT) is applied to the acceleration sequence for each time period (3 hours, totaling 1,080,000 data points) to transform it from the time domain to the frequency domain, obtaining the amplitude corresponding to each frequency component. Next, the obtained spectrum is divided according to a preset frequency band division standard. This standard is based on research into the acoustic emission characteristics of this type of rock mass (granite) during the failure process. The frequency bands are divided into: low frequency band (0-20Hz), mid frequency band (20-60Hz), and high frequency band (60-100Hz). Then, the energy of all frequency components within each frequency band is accumulated. The energy of a single frequency component is proportional to the square of its amplitude, defined here as the square of the amplitude. The frequency band energy is the sum of the energy values ​​of all frequency components within that band. For example, for the steady-state region, after FFT and calculation, the accumulated energies of the low, mid, and high frequency bands are respectively... , , (Unit is) The total energy during this period is the sum of these three factors, i.e. Finally, the frequency band energy distribution is established. This is calculated by dividing the energy of each frequency band by the total energy. For the steady-state region, the energy proportion of the low-frequency band is... The proportion of mid-frequency band is The proportion of high-frequency bands is Similarly, the same operation is performed on the data in the critical state region to obtain its frequency band energy distribution, for example, 75.2% in the low frequency band, 18.1% in the mid frequency band, and 6.7% in the high frequency band. Finally, the frequency band energy distribution data for the two time periods are obtained.

[0082] S302: Based on the frequency band energy proportion distribution, identify the direction of energy proportion change in adjacent time periods, calculate the position change of the spectral centroid in each time period, determine the consistency between the energy direction and the centroid direction in the time series, and establish a direction consistency sequence.

[0083] Based on the frequency band energy distribution, the migration pattern of energy between different time periods is further analyzed. This process is carried out within the steady-state and critical-state regions with smaller time steps (e.g., 10 minutes). First, the direction of energy proportion change between adjacent time periods (every 10 minutes) is identified. Taking 2024-01-20-15:00-15:10 (time period T1) and 15:10-15:20 (time period T2) as examples, the frequency band energy proportions of T1 and T2 are calculated. Assuming that the high-frequency energy proportion of T1 is 8.5% and that of T2 is 7.9%, the change in the high-frequency energy proportion is negative. At the same time, the low-frequency energy proportion of T1 is 74.1% and that of T2 is 75.2%, so the change in the low-frequency energy proportion is positive. Based on this, it is determined that the main direction of energy change between adjacent time periods is migration from high frequency to low frequency, and the energy direction is recorded as -1. Conversely, it is recorded as +1, and no obvious migration trend is recorded as 0. Next, the spectral centroid position of each time period (T1 and T2) is calculated. The centroid of the spectrum is calculated by multiplying each frequency value by its corresponding energy (square of the amplitude), summing the results, and then dividing by the total energy. Assuming the calculated centroid of the spectrum for T1 is 25.5Hz and for T2 it is 23.8Hz, then the change in the centroid position is: Hz, with a negative direction of change, is denoted as -1 for the center of gravity direction. Finally, the consistency between the energy direction and the center of gravity direction in the time series is determined. In this example, the energy direction is -1, and the center of gravity direction is also -1; they are consistent, so the consistency index for this segment is +1. If they are opposite (e.g., energy migrates to higher frequencies, but the center of gravity moves to lower frequencies), it is denoted as -1. If either direction is 0, it is denoted as 0. This process is repeated for all consecutive 10-minute segments within the steady-state and critical-state regions, ultimately establishing a time series composed of +1, -1, and 0, i.e., the direction consistency sequence.

[0084] S303: Call the directional consistency sequence, summarize the directional differences in the steady-state and critical-state segments respectively, organize the difference index of coupling directions, and generate the energy centroid coupling difference surface;

[0085] A directional consistency sequence is invoked, consisting of a series of time-varying values ​​of +1 (consistent), -1 (inconsistent), and 0 (neutral). This sequence is divided into a steady-state region (2024-01-2012:00-15:00) and a critical-state region (2024-01-2015:00-18:00) as defined in S203. The steady-state region, totaling 3 hours, is further divided into 18 analysis periods at 10-minute intervals, yielding 17 consistency indices. The critical-state region also yields 17 consistency indices. Next, these directional difference indices are summarized within both the steady-state and critical-state regions. The summarization method involves calculating the "coupling directional difference index" for each region. This index is defined as the number of times the consistency index is +1 minus the number of times it is -1, divided by the total number of non-zero indices. The index ranges from -1 to 1. A value closer to +1 indicates a stronger coupling (consistency) between the energy migration direction and the direction of spectral centroid shift. For the steady-state region, assuming that out of its 17 consistency indices, 5 are +1, 4 are -1, and 8 are 0, then its coupling direction difference index is... For the critical state region, assuming that out of its 17 consistency indices, 12 are +1, 1 is -1, and 4 are 0, then its coupling direction difference index is: Finally, these two indices (0.11 and 0.85) are combined to form a difference index for the coupling direction. These two values ​​together constitute the energy centroid coupling difference surface. This difference surface reveals through specific numerical values ​​that in the steady-state region, the correlation between energy migration and the direction of movement of the spectral centroid is weak, exhibiting characteristics of random fluctuations; while in the critical state region, the two exhibit a very strong positive correlation, that is, energy systematically migrates to lower frequencies, and the spectral centroid also moves steadily to lower frequencies accordingly.

[0086] Please see Figure 5 The specific steps for obtaining load trigger comparison information are as follows:

[0087] S401: Based on the energy center of gravity coupling difference surface, analyze the peak ground motion sequence of the corresponding period, judge according to the triggering benchmark, count the occurrence frequency of ground ground triggering events in each segment, and generate earthquake triggering frequency data.

[0088] Based on the different rock mass states (steady state and critical state) indicated by the energy center-of-gravity coupling difference surface, the analysis stage of response to external loads is initiated. A triggering criterion is used to distinguish whether a seismic motion constitutes a triggering event. The triggering criterion can be selected as a preset peak ground acceleration (PGA) threshold, for example, determined based on the design seismic motion parameters of the target area or the correlation analysis between historical seismic motions and unstable rock mass events. First, the PGA sequence from January 1 to February 29, 2024, is obtained from the seismic monitoring network in the area where the unstable rock mass GTC-01 is located. This sequence records the daily maximum PGA value in g (gravitational acceleration). Next, a seismic triggering criterion is established. This criterion is set with reference to the design value of the bedrock horizontal PGA with a 50-year exceedance probability of 10% in the "China Seismic Ground Motion Parameter Zoning Map" (GB18306-2015), which is 0.10g. Considering the high sensitivity of the unstable rock mass, the triggering criterion is set to 20% of this design value, i.e., The verification process for this value is as follows: Analysis of all microseismic events greater than magnitude 2.0 that occurred in the region over the past five years revealed that over 95% of these events generated PGA values ​​below 0.02g, while ground motions exceeding this value showed a weak temporal correlation with historical small rockfall events. Therefore, 0.02g was considered a sensitive threshold for effectively screening potentially impactful disturbances. Subsequently, the PGA peak sequence was segmented according to the triggering criteria. Here, a "week" was used as the statistical period, dividing the two-month data into eight periods. The daily PGA values ​​recorded within each weekly period were assessed, and the frequency of events with PGA values ​​exceeding 0.02g was statistically analyzed. For example, in the first week of January 2024, the PGA values ​​for the seven days were [0.005g, 0.008g, 0.025g, 0.011g, 0.031g, 0.015g, 0.009g], with two days having PGA values ​​exceeding 0.02g. Therefore, the frequency of seismic triggering events in that week was 2. This statistical analysis was repeated for all 8 weekly periods, ultimately generating a seismic triggering frequency data sequence containing 8 values.

[0089] S402: Based on earthquake triggering frequency data, obtain the sequence of instability events in the monitoring records of unstable rock masses during the same period, count the frequency of occurrence of instability events, perform proportional offset calculation on earthquake triggering frequency and instability event frequency, and obtain probability offset amplitude data;

[0090] Based on earthquake trigger frequency data, a sequence of unstable rock mass events during the same period (January 1 to February 29, 2024, a total of 8 weeks) was obtained from the daily inspection records and event logs of the automated video monitoring system of the GTC-01 unstable rock mass. An unstable event was defined here as any observable rock debris fall or small-scale collapse. Statistically, the frequency of unstable events recorded within the 8 weeks was times / week. Next, a proportional offset calculation was performed on the earthquake trigger frequency and unstable event frequency for each week. First, the "trigger-instability ratio" for each period was calculated, i.e. For example, the proportion in week five was... If the trigger frequency is 0 in a certain week, then that week will not be included in the calculation. Then, a baseline ratio is needed. This baseline proportion was obtained by analyzing data from the unstable rock mass during the period when it was assessed as "long-term stable" (e.g., the entire year of 2022). In 2022, there were a total of 50 triggering events and 2 instability events, hence the baseline proportion. Finally, the probability offset amplitude for each weekly period is calculated, which is the relative difference between the current period's proportion and the baseline proportion: Taking week five as an example, its probability shift amplitude is... For the first four weeks, since there were zero instability events, the proportion was 0, and the probability shift amplitude was [missing value]. For all weeks with triggering events, a probability offset amplitude data sequence is obtained, for example, [-1.0,-1.0,-1.0,-1.0,5.25,4.0,5.25,7.33].

[0091] S403: Based on the probability offset amplitude data, evaluate the overall offset intensity based on the probability offset amplitude of multiple periods, organize the relationship between corresponding load triggering and probability offset, and generate load triggering comparison information;

[0092] Based on the probability offset amplitude data sequence [-1.0, -1.0, -1.0, -1.0, 5.25, 4.0, 5.25, 7.33], a comprehensive evaluation of the probability offset amplitude across multiple periods is conducted to assess the overall offset intensity. Here, "multiple periods" refers to the eight weekly time periods defined in S402. The specific method for assessing the overall offset intensity is to treat the probability offset amplitude values ​​of these eight weeks as a time series and perform trend analysis. A linear regression method is used here, with time (week number, t=1 to 8) as the independent variable and the probability offset amplitude value as the dependent variable, to fit a straight line. The slope of the linear regression equation was obtained through calculation. This slope This is defined as the "overall offset strength". In this example, a linear regression is performed on the sequence [1,-1.0],[2,-1.0],…,[8,7.33] to obtain the slope. This positive slope indicates that, over time, under the same or similar seismic triggering levels, the probability of unstable rock masses exhibits a significant increasing trend. The next step is to organize the relationship between load triggering and probability shift, that is, to correlate the frequency of triggering events with the intensity of probability shift. Here, the 8-week data is divided into two groups according to triggering frequency: a low-triggering group (1-3 times / week, 4 weeks) and a high-triggering group (4-6 times / week, 4 weeks). The average probability shift amplitude is calculated for both groups: -1.0 for the low-triggering group and [missing value] for the high-triggering group. The final generated load trigger comparison information is a structured data set containing: {overall offset strength: 1.35, average offset of low trigger group: -1.0, average offset of high trigger group: 5.46}.

[0093] Please see Figure 6 The specific steps for obtaining the joint verification data for each stage are as follows:

[0094] S501: Based on load-triggered comparison information, according to the time series of crack widths in the same unstable rock mass, the deformation increment calculation is performed on the widths of adjacent time periods to obtain the crack propagation rate sequence, and the time points corresponding to the rate inflection points are selected to generate crack rate feature point data.

[0095] Based on the trend of rock mass condition changes revealed by load-triggered comparison information, width monitoring data of a major fracture on the GTC-01 unstable rock mass was introduced for final verification. Time series of fracture width data were obtained from January 1st to February 29th, 2024, every 6 hours, using multi-point displacement gauges installed on this fracture. First, deformation increment calculations were performed on the width values ​​of adjacent time periods, i.e. Where W is the crack width. Then, the deformation increment is divided by the time interval (6 hours) to obtain the crack propagation rate sequence. For example, if the width is measured to be 20.50 mm at 12:00 on a certain day and 20.53 mm at 18:00, then the expansion rate during that period is... mm / hour. This calculation is performed over the entire monitoring period to obtain a complete fracture propagation rate sequence. The next step is to filter out inflection points in this rate sequence. Inflection points are identified by calculating the rate of change of the rate, i.e., acceleration: When the acceleration of the speed When the sign of a rate changes (from positive to negative or from negative to positive), that point in time is identified as a rate inflection point.

[0096] Table 2. Data on fracture velocity characteristics.

[0097]

[0098] As shown in Table 2, by traversing the acceleration changes of the entire rate sequence, all rate inflection points that meet the conditions and their corresponding time points are selected. Finally, the time points of these inflection points are collected to generate fracture rate feature point data, such as [2024-01-2018:00, 2024-02-05 06:00, 2024-02-15 12:00].

[0099] S502: Call the fracture rate feature point data, extract the corresponding time points in the displacement sequence, and perform synchronous extraction on the displacement change and cumulative change for each time period to obtain the fracture and displacement variation data set.

[0100] The fracture rate feature point data, namely [2024-01-2018:00, 2024-02-05 06:00, 2024-02-15 12:00], is used to divide the entire monitoring period (January 1st to February 29th) into multiple stages. The boundaries of these stages are these feature points. For example, stage one is from the start of monitoring to the first inflection point, stage two is between the first and second inflection points, and so on. Then, the displacement values ​​at the start and end times of each stage are extracted from the original fracture width monitoring sequence (i.e., displacement sequence). For each stage, the displacement change and cumulative change are extracted simultaneously. The displacement change is defined as the cumulative displacement at the end of the stage minus the cumulative displacement at the start. The cumulative change is the total fracture width at the end of the stage.

[0101] Table 3 Data set of crack and displacement variation

[0102]

[0103] As shown in Table 3, taking stage 3 as an example, its start time is 06:00 on February 5, 2024, and its end time is 12:00 on February 15, 2024. At the beginning of this stage, the cumulative displacement is 29.1 mm, and at the end, it is 33.2 mm. Therefore, the displacement change in this stage is... mm. The cumulative change at the end of this stage was 33.2 mm. This process was repeated for all stages defined by the inflection point, resulting in a structured data set, namely the crack and displacement variation data set, which correlates different evolution stages with specific deformation quantification indicators.

[0104] S503: Based on the crack and displacement variation data set, the crack propagation rate, displacement change, cumulative change and load trigger comparison information are normalized to quantify the stage state strength of the unstable rock mass and generate stage joint verification data.

[0105] Based on the fracture and displacement variation data set, a final quantitative assessment of the unstable rock mass's condition was conducted. The last stage of the analysis was selected, namely stage 4 (2024-02-15-12:00 to 2024-02-29-23:59). Four core indicators for this stage were extracted from the relevant steps: 1) Average fracture propagation rate, calculated by dividing the stage displacement change by the stage duration. mm / hour; 2) Displacement change, mm; 3) Cumulative change, mm; 4) Load trigger comparison information, here using the overall offset strength calculated by S403. To integrate these four indicators with different dimensions and properties, normalization is required. The maximum value benchmark used for normalization is determined based on records of similar unstable rock masses before historical instability, or according to the limit state values ​​in design specifications. The maximum value benchmark set for this embodiment is as follows: maximum rate. mm / hour, maximum stage displacement change mm, maximum cumulative change mm, maximum load offset strength The determination of these maximum values ​​was verified through the following experiments: Complete monitoring data from 10 similar unstable and failed rock mass cases were selected, and various indicators in their last evolutionary stage before failure were statistically analyzed. The 95th percentile value was taken as the benchmark for the maximum value, ensuring the rationality and representativeness of the benchmark value. The normalized calculation is as follows: Finally, the overall strength of the unstable rock mass at this stage was quantified by weighted summation. The weighting coefficients were also determined based on the AHP expert scoring method, reflecting the contribution of each indicator to instability: rate weight. Change weight Cumulative weight Load response weights Phase state strength The final calculated value of 0.6128 is the stage joint verification data.

[0106] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for verifying the stability evaluation results of unstable rock masses based on historical data comparison, characterized in that, Includes the following steps: S1: Obtain the frequency monitoring sequence of the unstable rock mass, analyze the continuous changes over time, calculate the slope of the frequency change and obtain the slope difference, identify the period of significant dynamic change, divide the stable segment and the attenuation segment, divide the monitoring sequence into multiple dynamic behavior categories, and generate working condition slope change amplitude groups. S2: Based on the slope change range group of the working conditions, analyze the rock mass quality, structural surface dip angle, cohesion and internal friction angle sequence of the corresponding period, perform differential operation on the mechanical parameters to form an offset sequence, identify the offset inflection point and boundary period, construct the difference data between steady state and critical state, and generate the offset difference surface between steady state and critical state. S3: Based on the steady-state and critical-state offset difference surface, calculate the acceleration spectrum energy proportion distribution, determine the consistency between the spectrum centroid direction and the energy migration direction, summarize the directional differences within the steady-state and critical-state segments, and generate the energy centroid coupling difference surface; S4: Based on the energy centroid coupling difference surface, analyze the peak ground motion sequence, count the triggering frequency in segments according to the triggering benchmark, calculate the occurrence frequency in combination with the monitoring instability sequence, compare the proportional offset of the two types of frequencies and integrate them across periods to generate load triggering comparison information.

2. The method for verifying the stability evaluation results of unstable rock masses based on historical data comparison according to claim 1, characterized in that, The slope variation range group includes the slope range of the stable segment, the slope range of the attenuation segment, and the slope range of the transition segment. The steady-state and critical-state offset difference surface includes the steady-state parameter offset field, the critical-state parameter offset field, and the steady-state-critical-state transition offset gradient field. The energy centroid coupling difference surface includes the energy proportion variation field, the spectral centroid drift field, and the energy and centroid direction coupling coefficient distribution. The load triggering comparison information includes the seismic ground motion triggering frequency distribution, the instability event response frequency distribution, and the triggering probability offset distribution field.

3. The method for verifying the stability evaluation results of unstable rock masses based on historical data comparison according to claim 1, characterized in that, The specific steps for obtaining the slope variation amplitude group under the operating conditions are as follows: S101: Obtain the frequency monitoring sequence of the unstable rock mass, detect the frequency corresponding to each time period on the time axis and the time difference between adjacent time periods, calculate the frequency change slope of adjacent time periods, screen time slices with significant slope changes, and establish a dynamic change time slice group. S102: Based on the dynamic change time slice group, compare the slope sequence before and after the target time slice, calculate the slope change amplitude of each time period, determine the numerical range of the slope change amplitude, divide the stable segment and the decay segment, and generate a dynamic behavior interval group. S103: Based on the dynamic behavior interval group, collect the slope change amplitude of multiple time periods, classify and organize the dynamic behavior categories, and generate the working condition slope change amplitude group.

4. The method for verifying the stability evaluation results of unstable rock masses based on historical data comparison according to claim 3, characterized in that, The process of determining the numerical range of slope change and dividing the stable segment into the attenuation segment is as follows: Using the equal time intervals between adjacent dynamic change time slices and without dynamic change time slices as reference intervals, the median and dispersion of the slope change amplitude sequence within the reference intervals are calculated respectively, and the smaller of the medians of the two reference intervals is used as the stability benchmark of the slope change amplitude. The slope change amplitude sequence within the dynamic change time slice is used as the upper quantile, and the slope change amplitude stability benchmark is combined to set the slope change amplitude attenuation judgment threshold. The slope change amplitude sequence is compared with the slope change amplitude stability benchmark and the slope change amplitude attenuation judgment threshold in chronological order. When the slope change amplitude is between the slope change amplitude stability benchmark and the slope change amplitude attenuation judgment threshold, it is divided into a stable segment. When the slope change amplitude exceeds the slope change amplitude attenuation judgment threshold, it is divided into an attenuation segment.

5. The method for verifying the stability evaluation results of unstable rock masses based on historical data comparison according to claim 3, characterized in that, The specific steps for obtaining the steady-state and critical-state shift difference surface are as follows: S201: Based on the slope change range group of the working conditions, obtain the rock mass quality, structural surface dip angle, cohesion and internal friction angle sequence for the corresponding period, perform differential operation on each mechanical parameter at the same time point, and establish a multi-parameter offset sequence; S202: Based on the multi-parameter offset sequence, compare the composite change amplitude of each parameter, identify the offset inflection point and the corresponding boundary time period, compare the composite change amplitude before and after the time period, and obtain the offset boundary interval data. S203: Based on the offset boundary interval data, calculate the concentration and variation amplitude of the offset sequence before and after the boundary, construct the difference data between the steady state and the critical state, and generate the offset difference surface between the steady state and the critical state.

6. The method for verifying the stability evaluation results of unstable rock masses based on historical data comparison according to claim 5, characterized in that, The process of identifying the offset inflection point and the corresponding boundary time period is as follows: Using the time periods at the beginning and end of the time axis in the multi-parameter offset sequence that do not show abrupt changes in the composite change amplitude as reference segments, the median and dispersion of the composite change amplitude at each time point within the reference segments are calculated, and the reference range of the composite change amplitude is set based on the median and dispersion. The median of the composite change range of the multi-parameter offset sequence over the entire time range is used as the candidate baseline for the offset inflection point, and the candidate baseline for the offset inflection point is jointly determined with the reference range of the composite change range to determine the threshold for the offset inflection point. The composite change amplitude and the offset inflection point determination threshold of each time point are compared sequentially according to the multi-parameter offset sequence. When the composite change amplitude continuously crosses the offset inflection point determination threshold and maintains a preset duration on both sides of the offset inflection point determination threshold, the time point corresponding to the position where the composite change amplitude crosses the offset inflection point determination threshold is determined as the offset inflection point, and the time segments before and after the offset inflection point that meet the preset duration conditions are determined as the corresponding boundary time periods. The preset duration is the time length corresponding to three consecutive monitoring time points.

7. The method for verifying the stability evaluation results of unstable rock masses based on historical data comparison according to claim 5, characterized in that, The specific steps for obtaining the energy centroid coupling difference surface are as follows: S301: Based on the difference surface between steady state and critical state offset, obtain the acceleration monitoring data for the corresponding time period, perform spectral transformation on the acceleration sequence for each time period, divide the spectrum into multiple frequency bands, and perform energy accumulation calculation on each frequency band to establish the frequency band energy proportion distribution; S302: Based on the energy proportion distribution of the frequency band, identify the direction of energy proportion change in adjacent time periods, calculate the position change of the centroid of the spectrum in each time period, determine the consistency between the energy direction and the centroid direction in the time series, and establish a direction consistency sequence. S303: Call the aforementioned direction consistency sequence, summarize the direction differences in the steady-state segment and the critical state segment respectively, organize the difference index of the coupling direction, and generate the energy centroid coupling difference surface.

8. The method for verifying the stability evaluation results of unstable rock masses based on historical data comparison according to claim 7, characterized in that, The specific steps for obtaining the load trigger comparison information are as follows: S401: Based on the energy centroid coupling difference surface, analyze the peak ground motion sequence of the corresponding period, judge according to the triggering benchmark, count the occurrence frequency of ground motion triggering events in each segment, and generate earthquake triggering frequency data. S402: Based on the earthquake triggering frequency data, obtain the instability event sequence in the monitoring records of unstable rock masses during the same period, count the occurrence frequency of instability events, perform proportional offset calculation on the earthquake triggering frequency and the instability event frequency, and obtain probability offset amplitude data; S403: Based on the probability offset amplitude data, evaluate the overall offset intensity based on the probability offset amplitude of multiple periods, organize the relationship between corresponding load triggering and probability offset, and generate load triggering comparison information.

9. The method for verifying the stability evaluation results of unstable rock masses based on historical data comparison according to claim 1, characterized in that, The method further includes: S5: Based on the load trigger comparison information, calculate the deformation increment and expansion rate of the same unstable rock mass crack width sequence, filter the time point corresponding to the rate inflection point, locate the corresponding time of the displacement sequence and take the displacement change and cumulative change, combine normalization processing, quantify the stage state strength of the unstable rock mass, and generate stage joint verification data. The stage joint verification data includes a crack propagation rate sequence, a displacement variation sequence, and a stage state intensity quantification sequence.

10. The method for verifying the stability evaluation results of unstable rock masses based on historical data comparison according to claim 9, characterized in that, The specific steps for obtaining the joint verification data at each stage are as follows: S501: Based on the load triggering comparison information, according to the time series of the same unstable rock mass crack width, perform deformation increment calculation on the width of adjacent time periods to obtain the crack propagation rate sequence, filter the time points corresponding to the rate inflection points, and generate crack rate feature point data. S502: Call the fracture rate feature point data, extract the corresponding time points in the displacement sequence, and perform synchronous extraction on the displacement change and cumulative change for each time period to obtain a fracture and displacement variation data set. S503: Based on the fracture and displacement variation data set, the fracture propagation rate, displacement change, cumulative change and load trigger comparison information are normalized to quantify the stage state strength of the unstable rock mass and generate stage joint verification data.

Images