Coupling method of slope multi-scale sliding index combining macroscopic crack and microscopic displacement

Abstract

The application discloses a slope multi-scale sliding index coupling method fusing macroscopic cracks and microscopic displacement, and relates to the technical field of slope safety monitoring and early warning. In view of the problem that the existing multi-source fusion simply weights and superimposes macroscopic cracks and microscopic displacement, and early warning is difficult to consider early and accurate, the method collects the macroscopic crack opening amount and the microscopic displacement amount of the same monitoring object and aligns them, calculates the macroscopic scale index and the microscopic scale index normalized by the relative baseline fluctuation band in the sliding time window, calculates the cross-scale consistency coefficient of the two, and uses the cross-scale consistency coefficient as a gating factor to modulate the microscopic scale index. When the consistency is not less than the confirmation threshold, the microscopic scale index is enhanced; when the consistency is less than the confirmation threshold, the microscopic scale index is inhibited, so that the cross-scale coupling sliding index is obtained. The cross-scale coupling sliding index is compared with the adaptive grading threshold to determine the sliding grade and output the early warning. The macroscopic crack synchronously confirms and suppresses the environmental noise false alarm, and the microscopic displacement is sensitive to the advance amount, so that the method is early, accurate, and light in calculation, and can be used for slope sliding early warning of hydropower, mines and the like.

Description

Technical Field

[0001] This invention relates to the field of slope safety monitoring and early warning technology, and in particular to a method for coupling multi-scale slippage indices of slopes that integrates macroscopic cracks and microscopic displacements. Background Technology

[0002] Landslides caused by slope instability are major geological hazards in the construction and operation of hydropower, mining, and transportation projects. The core of slope landslide early warning is to seize the critical moment when the slope transitions from relative stability to accelerated instability, allowing sufficient time for early warning and response before the entire slope slides down. Early slope safety management relied heavily on manual inspections and single-point displacement observations, which could only be detected passively after deformation had become obvious, making it difficult to provide reliable early warnings.

[0003] To detect slope deformation in advance, the current mainstream approach is to deploy a multi-source monitoring network on the slope. This involves simultaneously deploying sensors such as crack gauges, extensometers, displacement gauges, and GPS receivers on the slope surface and in the shallow layers. This acquires both macroscopic crack opening data reflecting the development of surface cracks and microscopic displacement data reflecting the slow creep of the slope. Based on this, there are single displacement criteria such as the inverse velocity method based on the inverse of displacement rate and the improved tangent angle method based on the tangent angle of the displacement-time curve. There are also multi-source fusion methods that integrate multiple monitoring indicators into a comprehensive early warning index using weighted averaging, Bayesian estimation, or evidence theory, and then compare it with a pre-set threshold.

[0004] However, macroscopic crack opening and microscopic displacement are two signals with different characteristics. Microscopic displacement is like body temperature, reacting sensitively and changing early, but it is easily inflated by reversible environmental disturbances such as diurnal temperature differences, expansion and contraction caused by rainfall infiltration, and freeze-thaw cycles. After a heavy rainfall, slope displacement readings generally jump, but this does not necessarily mean that the slope is actually on the verge of landslides. Macroscopic crack opening, on the other hand, is like the tearing of a wound, reflecting the irreversible accumulation of damage to the soil and rock mass, with fewer false alarms. However, by the time the main control cracks have clearly accelerated their opening, there is very little lead time left for evacuation. The existing weighted fusion method directly adds the two signals with fixed weights, which is equivalent to tying a sensitive but erratic pointer and a reliable but sluggish pointer together and averaging them. This method cannot fully utilize the lead time provided by the sensitive pointer, nor can it suppress the frequent false alarms caused by the sensitive pointer.

[0005] The root cause lies in the fact that existing fusion methods only concern themselves with the magnitudes of macroscopic and microscopic signals, neglecting whether a jump in microscopic displacement at a given timeframe is confirmed by the expansion of macroscopic cracks during the same period. A jump in microscopic displacement caused by reversible environmental disturbances is isolated; it does not accompany the synchronous opening of macroscopic cracks within the same timeframe. True precursors to landslides, however, are characterized by a synchronous acceleration of microscopic displacement and macroscopic cracks over time, with the ratio of their increments falling within a range consistent with the slope's deformation pattern. The theory of progressive failure of soil and rock masses and instability of locked sections has long demonstrated that the closer a slope approaches a landslide-prone state, the more co-developing its macroscopic and microscopic deformations become. This observed objective phenomenon has consistently failed to be used by existing fusion methods as an engineering criterion for distinguishing between environmental noise and true precursors.

[0006] Therefore, how to achieve early and accurate early warning of slope slippage without introducing complex models and extensive training, by leveraging the sensitivity of micro-displacement to gain early warning time and by using the confirmatory effect of macro-cracks to suppress false alarms caused by environmental noise, is a technical problem that urgently needs to be solved in this field. Summary of the Invention

[0007] To address the problem in existing multi-source fusion early warning technologies for slopes that simply weight and superimpose macroscopic cracks and microscopic displacements, making it difficult to balance early warning with accuracy, this invention provides a multi-scale slope slip index coupling method that integrates macroscopic cracks and microscopic displacements. This method uses microscopic displacement to drive early response while simultaneously suppressing false alarms due to environmental noise through synchronous confirmation of macroscopic cracks, thus achieving both early and accurate slope slip warnings with lightweight window-by-window calculations. After repeatedly comparing long-term monitoring data from multiple sedimentary layers and rock slopes, the inventors discovered that true slip precursors are always accompanied by the synchronous acceleration of macroscopic cracks and microscopic displacements, while isolated microscopic upward jumps caused by rainfall or temperature differences lack this cross-scale synchronicity. This pattern has not previously been systematically used as an engineering criterion to distinguish between noise and precursors.

[0008] The method of this invention includes the following steps: S1, during the initial run, the macroscopic crack opening amount and microscopic displacement amount of the slope during its stable period are obtained, and the confirmation threshold and grading threshold are determined accordingly. Subsequently, the macroscopic crack opening amount and microscopic displacement amount of the same monitoring object on the slope are collected and aligned according to a unified time reference to obtain a macroscopic crack sequence and a microscopic displacement sequence; S2, within a sliding time window, the crack propagation rate of the macroscopic crack sequence and the displacement rate of the microscopic displacement sequence are calculated respectively, and the crack propagation rate and the displacement rate are normalized relative to their baseline fluctuation bands within a preset retrospective period. S3. Within the sliding time window, calculate the cross-scale consistency coefficient of the macro-scale index and the micro-scale index, and use the cross-scale consistency coefficient as a gating factor to modulate the micro-scale index. If the cross-scale consistency coefficient is not less than the confirmation threshold, perform confirmation enhancement; otherwise, perform suppression to obtain a cross-scale coupled slippage index. S4. Compare the cross-scale coupled slippage index with the classification threshold to determine the slippage level. S5. Output early warning information or trigger corresponding handling actions according to the slippage level.

[0009] This invention also provides a slope multi-scale slippage index coupling system corresponding to the above method, integrating macroscopic cracks and microscopic displacements. The system includes a data acquisition unit, a dual-scale index unit, a cross-scale coupling unit, a hierarchical judgment unit, and an early warning output unit. Each unit respectively performs the functions of data acquisition alignment, dual-scale normalized index calculation, cross-scale consistency gating coupling, hierarchical judgment, and early warning output in the above method. This invention also provides a computer-readable storage medium and an electronic device, wherein a computer program thereon, when executed by a processor, implements the steps of the above method.

[0010] The beneficial effects of this invention are as follows: First, by using the cross-scale consistency coefficient of macro-scale and micro-scale indicators within the same sliding time window as a gating factor, sensitive micro-scale indicators are confirmed, enhanced, or suppressed. This suppresses isolated micro-displacement jumps caused by environmental disturbances that are not confirmed by macro-cracks to low values, thus preventing false alarms. Meanwhile, micro-deformations that are truly confirmed synchronously by macro-cracks are enhanced to high values, triggering early warnings. This fundamentally resolves the contradiction between early and accurate predictions in existing weighted superposition methods. Second, the rate of each scale is normalized relative to its own baseline fluctuation band, and the preset consistency band is calibrated by the deformation mode of this slope. This makes two signals with vastly different dimensions and magnitudes comparable, and provides a clear physical basis for consistency discrimination. Third, both the grading threshold and the confirmation threshold are adaptively calibrated by the historical baseline of the slope during its stable period, eliminating the need for manual recalibration across slopes. Fourth, the entire coupling discrimination is calculated at a constant level window by window, without relying on neural networks or training, facilitating real-time operation on the monitoring terminal. Attached Figure Description

[0011] Figure 1 This is a schematic diagram of the overall process of the method described in this invention; Figure 2 This is a schematic diagram of the module structure of the multi-scale slope slip index coupling system described in this invention; Figure 3 This is a schematic diagram illustrating the principle of cross-scale consistency gating fusion as described in this invention; Figure 4 This is a schematic diagram of the graded threshold axis of the cross-scale coupled slippage index described in this invention; Figure 5 This is a schematic diagram illustrating the decision-making process of cross-scale consistency gating as described in this invention; Figure 6 This is a schematic diagram of the sub-process of the cross-scale consistency gating coupled computation described in this invention; Figure 7 This is a schematic diagram comparing the false alarm rates of the present invention and existing methods during periods of rainfall and noise. Figure 8 This is a schematic diagram illustrating the evolution of cross-scale coupled slip index and cross-scale consistency coefficient during the slope approaching slip process in one embodiment of the present invention. Figure 9 This is a schematic diagram illustrating how the cross-scale consistency coefficient changes with the degree of synchronization increase and the degree of consistency band conformity as described in this invention. Figure 10 This is a schematic diagram of the structure of the electronic device described in this invention. Detailed Implementation

[0012] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. The specific sampling intervals, time window lengths, backtracking periods, threshold values, congruence zone boundaries, normalization factors, and various monitoring values ​​used in the following embodiments are all illustrative examples and do not constitute a limitation on the scope of protection of this invention. Those skilled in the art can adjust the corresponding parameters according to the slope soil and rock conditions, the accuracy of the monitoring instruments, and the actual working conditions; the adjusted embodiments still fall within the scope of protection of this invention.

[0013] The monitored object refers to a local slope section that requires focused monitoring, typically a main controlling crack and the slope sections on both sides, or a slope unit covered by the same monitoring pier. The key to this invention is deploying sensors that measure the macroscopic crack opening and sensors that measure the microscopic displacement at the same monitored object, ensuring that observations at both macroscopic and microscopic scales fall in the same spatial location, thus enabling mutual verification and comparability.

[0014] The macroscopic crack opening refers to the width between the two walls of the visible main control crack on the slope surface, measured in millimeters. It is directly measured by crack gauges or extensometers installed across the crack and reflects the tensile failure development of the slope on a macroscopic scale. The microscopic displacement refers to the cumulative displacement of the monitored object relative to the distant stability benchmark, measured in millimeters. It is measured by displacement gauges or global navigation satellite system receivers. Its magnitude in the early stage of landslides is usually much smaller than the change in crack opening and it is more sensitive to environmental disturbances; therefore, it is called the microscopic scale relative to crack propagation. The macroscopic crack sequence refers to the numerical sequence of macroscopic crack openings arranged according to the sampling time. The microscopic displacement sequence refers to the numerical sequence of microscopic displacements arranged according to the sampling time.

[0015] The unified time reference refers to normalizing the sampling times of the two sensors to the same clock and resampling to a common sampling interval time scale, so that macroscopic and microscopic observations within the same window correspond one-to-one in time. The sliding time window refers to a time interval of fixed length that slides forward continuously with the current time, for example, a length of 24 hours, sliding forward once every 1 hour. All calculations in this invention are completed within the current sliding time window. The sliding step size is usually taken as 1 / 24 to 1 / 12 of the sliding time window length to ensure a balance between time resolution and computational efficiency; those skilled in the art can adjust it within this range according to the sampling frequency and early warning response requirements.

[0016] The crack propagation rate refers to the rate of change of macroscopic crack opening with respect to time within a sliding time window, i.e., the net increase in crack opening within the window divided by the window length; the displacement rate refers to the rate of change of microscopic displacement with respect to time within the same sliding time window. The preset backtracking period refers to a relatively long historical interval preceding the current moment, used to characterize the normal fluctuation level at that location, for example, the most recent 30 days; the length of the preset backtracking period should be at least 10 times the length of the sliding time window, and not less than the complete cycle of the main environmental disturbance cycle of the slope (such as the rainfall cycle, freeze-thaw cycle). Those skilled in the art can determine the length of the backtracking period based on the specific working conditions of the slope under this principle. The baseline fluctuation band refers to the range of normal fluctuations in the crack propagation rate or displacement rate within the preset backtracking period due to measurement noise and reversible environmental disturbances. This invention uses the median of this rate within the preset backtracking period to characterize its normal level, and the absolute deviation of the median to characterize its normal fluctuation amplitude; both jointly define the baseline fluctuation band. The absolute deviation of the median refers to the median of the absolute values ​​of the differences between each sample and its median, which is more robust to occasional jumps than the standard deviation.

[0017] The normalization refers to converting the dimensional rate into a dimensionless relative intensity scaled by the baseline fluctuation band. This is achieved by subtracting the corresponding median from the current rate and then dividing by the product of a preset multiple and the absolute deviation of the median. The preset multiple ranges from 3 to 6, typically 4. Its physical meaning is that only rates exceeding normal fluctuations by several times are considered significant increases, thus suppressing the misreporting of normal fluctuations as distortions. The specific value of the preset multiple is determined based on the perturbation index of each rate: the perturbation index η = MAD / |median| is defined, where MAD is the absolute deviation of the median rate, and |median| is the absolute value of the median rate. When |median| is less than 0.001, the perturbation is considered extremely large, and the preset multiple is directly set to 6. Otherwise, when η > 0.3, the perturbation is considered large, and the preset multiple is set to 6; when η < 0.1, the perturbation is considered small, and the preset multiple is set to 3; when it falls between these two values, linear interpolation is used, or the default value of 4 is taken. For crack propagation rate and displacement rate, their respective disturbance indices are calculated independently and preset multiples are determined. The macro-scale index refers to the dimensionless quantity obtained after normalizing the crack propagation rate; the micro-scale index refers to the dimensionless quantity obtained after normalizing the displacement rate. Both are greater than 0, indicating that the deformation rate of the corresponding scale significantly exceeds its normal fluctuation and develops upward, and the larger the value, the further it deviates from the normal level.

[0018] The cross-scale consistency coefficient is the core quantity defined in this invention. It refers to a comprehensive measure of whether macro-scale indicators and micro-scale indicators rise synchronously and in the same direction within the same sliding time window, and whether the ratio of their increments falls within a preset consistency zone that is commensurate with the deformation mode of this slope. The value range is not less than 0 and not greater than 1. The closer it is to 1, the more mutually confirmed the deformation of the two scales is, and the more likely it is to characterize the true precursor. The closer it is to 0, the more isolated the rise of a certain scale is, and the more likely it is to characterize environmental noise. The synchronous upward trend refers to the simultaneous presence of macro-scale and micro-scale indicators greater than 0 within the same sliding time window, and both showing an upward trend within the window. The upward trend within the sliding time window is determined as follows: Let the sequence values ​​within the sliding time window be x1, x2, ..., xm, where m is the number of data points within the window. If both of the following conditions are met simultaneously, it is considered an upward trend: (a) the ending value is greater than the starting value, i.e., xm > x1; (b) the linear slope k fitted using the least squares method is greater than 0, and the p-value of the slope significance test is less than 0.1. For slope slip warnings, to reduce the false negative rate (avoiding missing true precursors), a significance level of 0.1 is used; those skilled in the art can adjust this threshold within the range of 0.05 to 0.2 according to the engineering risk level and the sensitivity to false alarms / false negatives. If any of the above conditions are not met, it is considered a non-upward trend. The degree of synchronous ascent refers to the dimensionless quantity that is the ratio of the smaller of the two ascents to the larger of the two ascents are established, and otherwise is 0, used to characterize the degree of equilibrium and synchronization of the two-scale ascents.

[0019] The increment ratio refers to the ratio of macro-scale indicators to micro-scale indicators; the preset consistency zone refers to an increment ratio interval defined by a lower and upper bound, where the increment ratio falling within this zone indicates that the relative proportion of macro-scale and micro-scale deformations is consistent with the deformation pattern of this slope; the lower bound L and upper bound U of the preset consistency zone are uniformly calibrated by the following method: collecting all sliding time windows that satisfy "both macro-scale indicators and micro-scale indicators are greater than 0" during the stable period of this slope, and calculating the increment ratio R_i in each effective window = macro-scale indicator_i / micro-scale indicator_i. The degree index _i is used to obtain the increment ratio sequence {R1,R2,…,Rn}. The corresponding coverage probability is selected based on the deformation mode of the slope: for traction slopes, the P20 quantile is used as the lower bound L and the P80 quantile as the upper bound U (covering 60% of the data); for shoving slopes, the P30 quantile is used as the lower bound L and the P70 quantile as the upper bound U (covering 40% of the data); if the deformation mode is unknown or mixed, the P10 quantile is used as L and the P90 quantile as U (covering 80% of the data). A higher coverage probability results in a wider consistency band and a higher tolerance for increment ratio deviations. Those skilled in the art can adjust the quantiles according to the dispersion of historical slope sliding data under this principle. The consistency band conformity refers to a dimensionless quantity where the increment ratio is 1 when it falls within the preset consistency band and monotonically decreases to 0 as it deviates from the interval when it falls outside the interval. The consistency band conformance is set to 1 when the increment ratio r belongs to the closed interval [L, U] (inclusive of the boundary); when r falls outside the interval, it is calculated using a linear decay formula: consistency band conformance = max(0, 1 - |rc| / d), where c = (L + U) / 2 is the center of the consistency band, d = (UL) / 2 is the half-width of the consistency band, and |rc| is the absolute distance from the center. The cross-scale consistency coefficient is the product of the synchronous rise degree and the consistency band conformance.

[0020] The gating factor refers to the modulation factor that uses the cross-scale consistency coefficient as a gate to determine whether sensitive micro-scale indicators are amplified or suppressed. The confirmation threshold is the threshold for determining whether the consistency is sufficient to constitute confirmation; its value is not less than 0.5 and not greater than 0.8, determined by the P80 quantile of the cross-scale consistency coefficient distribution during the slope's stable period. If the calculated quantile is less than 0.5, then 0.5 is used; if it is greater than 0.8, then 0.8 is used; otherwise, the calculated value is used. It should be noted that if the P80 quantile of the consistency coefficient distribution during the stable period is generally lower than 0.5, it indicates that the natural synchronicity between macro and micro signals of the slope is weak, and this method may not be applicable. It is necessary to adjust it based on engineering experience or adopt other early warning strategies. The confirmation enhancement refers to multiplying the microscale index by an enhancement coefficient of not less than 1, which increases with the consistency coefficient, when the cross-scale consistency coefficient is not less than the confirmation threshold. The confirmation enhancement is achieved by taking the cross-scale coupled slippage index as the product of the microscale index and the enhancement coefficient, where the enhancement coefficient α = 1 + γ, and γ is the cross-scale consistency coefficient. The suppression refers to multiplying the microscale index by a suppression coefficient less than 1, which decreases to 0 as the consistency coefficient decreases, when the cross-scale consistency coefficient is less than the confirmation threshold. The enhancement coefficient is the coefficient of not less than 1 multiplied by the confirmation enhancement; the suppression coefficient is the coefficient of less than 1 multiplied by the suppression. The suppression is achieved by taking the cross-scale coupled slippage index as the product of the microscale index and the suppression coefficient, where the suppression coefficient β = γ, which is equal to the cross-scale consistency coefficient itself.

[0021] The cross-scale coupled slip index refers to the final slip discrimination quantity obtained after gating and modulating the micro-scale index as described above. The slip level refers to the degree of urgency of slope slippage classified according to the magnitude of the cross-scale coupled slip index. This invention divides it into stable level, initial movement level, and slippage level. The classification threshold includes a first classification threshold and a second classification threshold. The first classification threshold refers to the threshold for determining whether a slope has progressed from the stable level to the initial movement level. The second classification threshold refers to the threshold for determining whether a slope has progressed from the initial movement level to the slippage level. Both are adaptively calibrated by the mean and standard deviation of the cross-scale coupled slip index during the stable period of this slope, using a preset multiple. The historical baseline refers to the normal level and dispersion of the corresponding quantity during the stable period of this slope. The deformation mode refers to the mechanical development pattern of slope instability, including the traction type where the slope toe pulls the upper slope body first, and the pushing type where the upper slope body pushes the lower slope body first. Different deformation modes correspond to different deformation ratios between macroscopic crack expansion and microscopic displacement, and are therefore used to calibrate the boundary of the preset consistency zone. The term "stable level" refers to a level where the slope is in a state of normal fluctuation; the term "initial movement level" refers to a level where the slope has shown confirmed early deformation and requires close attention; and the term "near-slip level" refers to a level where the slope is close to overall sliding and requires immediate action.

[0022] The following is combined with Figure 1 The overall process of the method described in this invention will be explained in detail. For example... Figure 1 As shown, the method of the present invention runs at a fixed rate on the monitoring terminal or monitoring platform. Each time the sliding time window slides forward once, S1 to S5 are executed sequentially to obtain the current sliding level and corresponding output.

[0023] S1 is for initialization and data acquisition alignment. During the initial system run, the macroscopic crack opening and microscopic displacement data for a period during the slope's stable phase are acquired to calibrate the confirmation and grading thresholds, serving as the basis for subsequent judgments. The specific calibration methods for the confirmation and grading thresholds are explained in the following sections on S3 and S4. Subsequently, the following acquisition and alignment are performed within each sliding time window: Crack gauges or extensometers are installed across the main control crack at the monitored object to collect macroscopic crack opening data, and displacement gauges or GPS receivers are installed at the same monitored object to collect microscopic displacement data. Since the sampling times and frequencies of the two sensors are usually inconsistent, they are unified to the same clock and resampled to a common sampling interval. Missing data points are filled with the effective values ​​of the nearest neighbor time, thus obtaining a time-corresponding macroscopic crack sequence and microscopic displacement sequence. This step provides a time-aligned, spatially co-located dual-scale data foundation for subsequent cross-scale comparisons.

[0024] S2 is used for calculating the dual-scale normalized index. Within the current sliding time window, the crack propagation rate of the macroscopic crack sequence is calculated, i.e., the net increase in crack opening within the window is divided by the window length; the displacement rate of the microscopic displacement sequence is calculated, i.e., the net increase in displacement within the window is divided by the window length. Subsequently, the median and median absolute deviation of the crack propagation rate and displacement rate are statistically analyzed within a preset backtracking period. The median represents the normal level of the rate, and the median absolute deviation represents its normal fluctuation. The two define the baseline fluctuation band of the rate. Then, normalization is completed by subtracting the corresponding median from the current rate, dividing by the preset multiple, and the product of the median absolute deviation, to obtain the dimensionless macroscopic and microscopic scale indices. The median and median absolute deviation are used instead of the mean and standard deviation to make the baseline statistics robust to occasional jumps; the preset multiple is determined according to the aforementioned disturbance index method, typically 4, and can be up to 6 when the slope background disturbance is large and up to 3 when the disturbance is small. Through this step, crack signals and displacement signals with vastly different dimensions and magnitudes are unified into dimensionless strength based on their respective normal fluctuations, thus enabling them to be compared on the same scale.

[0025] S3 is cross-scale consistency gating coupling, which is the core step of this invention. The principle of its cross-scale consistency gating fusion is as follows: Figure 3 As shown, the logic for gating is as follows: Figure 5 As shown, the calculation sub-process is as follows: Figure 6As shown, the process includes five sub-steps: calculating the degree of synchronous ascent (S31), calculating the degree of consistency band alignment (S32), multiplying the two to obtain the cross-scale consistency coefficient (S33), performing confirmation enhancement when the consistency coefficient is not less than the confirmation threshold (S34), and otherwise suppressing and outputting the cross-scale coupling slippage index (S35). This step first calculates the cross-scale consistency coefficient, then uses it to gate and modulate the microscale index. The cross-scale consistency coefficient is the product of the degree of synchronous ascent and the degree of consistency band alignment. The degree of synchronous ascent is calculated as follows: within the current sliding time window, if both the macroscale index and the microscale index are greater than 0, and both show an upward trend within the window according to the aforementioned upward trend determination method, then the ratio of the smaller of the two to the larger one is taken as the degree of synchronous ascent. The closer this ratio is to 1, the more balanced and synchronous the ascent of the two scales; the closer it is to 0, the weaker the ascent of one scale is compared to the other. If neither is simultaneously greater than 0, or neither shows an upward trend, then the degree of synchronous ascent is 0, indicating that there is no cross-scale synchronous deformation at this moment. The calculation of the consistency zone conformity is as follows: the ratio of the macro-scale index to the micro-scale index is taken as the increment ratio. When the increment ratio falls within the preset closed interval [L,U] of the consistency zone, it is taken as 1; when it falls outside the interval, it is calculated according to the aforementioned linear decay formula. The lower and upper boundaries of the preset consistency zone are marked according to the aforementioned quantile method. For example, for a traction slope where the toe moves first and the surface cracks lag behind, its consistency zone can be taken from the P20 / P80 quantile interval; for a lateral slope where the upper part moves first and the cracks open relatively ahead, its consistency zone can be taken from the P30 / P70 quantile interval. The cross-scale consistency coefficient changes with the degree of synchronous rise and the consistency zone conformity as follows: Figure 9 As shown, the cross-scale consistency coefficient is high only when both are high, thus reflecting the mutual confirmation of macroscopic and microscopic deformation.

[0026] After obtaining the cross-scale consistency coefficient, this step uses it as a gating factor to modulate the microscale index: when the cross-scale consistency coefficient is not less than the confirmation threshold, it is determined that the increase in microscale displacement has been confirmed by the synchronous propagation of macroscopic cracks. Therefore, the microscale index is enhanced by multiplying it by the enhancement coefficient α = 1 + γ (γ is the cross-scale consistency coefficient), thus amplifying the confirmed deformation. When the cross-scale consistency coefficient is less than the confirmation threshold, it is determined that the increase in microscale displacement is isolated and not confirmed, and is mostly caused by environmental disturbances. Therefore, the microscale index is suppressed by multiplying it by the suppression coefficient β = γ, thus suppressing the unconfirmed upward jump. The result after gating modulation is the cross-scale coupled slip index. Therefore, an isolated microscopic displacement spike caused by environmental disturbances such as rainfall, without synchronous opening of macroscopic cracks, is suppressed to a low value due to its very low cross-scale consistency coefficient and will not trigger an early warning. Conversely, a microscopic deformation synchronously confirmed by macroscopic cracks is amplified to a high value due to its high consistency coefficient, thus reflecting the precursory signs of slip earlier and more reliably than viewing either signal alone. One abnormality handling method in this step is as follows: when a sensor exhibits any of the following conditions within the window, it is considered a significant fault: (a) the number of consecutive missing measurement points exceeds 20% of the total number of points within the window; (b) a single data point exceeds the sensor's range; (c) the instantaneous rate change exceeds 10 times the absolute deviation of the median over its retrospective period. When any of the above faults occur, the cross-scale consistency coefficient is set to 0, and only the available scale indicators are output conservatively in a suppressed manner to avoid false amplification due to single-channel anomalies.

[0027] S4 is the grading determination, another key step in this invention. This step compares the cross-scale coupled slip threshold with the grading threshold to determine the slip level. The grading threshold axis and the division of the three slip levels—stable level, initial slip level, and slip level—are as follows: Figure 4As shown. The grading thresholds include a first grading threshold and a second grading threshold, both adaptively calibrated from the historical baseline of the slope's stable period: the mean and standard deviation of the cross-scale coupled slippage index during the slope's stable period are taken; the first grading threshold is the mean plus a smaller preset multiple of the standard deviation, and the second grading threshold is the mean plus a larger preset multiple of the standard deviation. The smaller preset multiple is typically 3, and the larger preset multiple is typically 8. During the judgment, if the cross-scale coupled slippage index is less than the first grading threshold, it is classified as stable; if it is not less than the first grading threshold and less than the second grading threshold, it is classified as initially moving; and if it is not less than the second grading threshold, it is classified as slippage-prone. Since the first and second grading thresholds are calibrated from the statistics of the slope's own stable period, the same judgment logic can be directly transferred to slopes with different soil and rock conditions and different instrument accuracies without the need for manual readjustment of the thresholds for each slope. To further suppress occasional fluctuations, this step requires that the slippage level be consistent within a continuous sliding time window before upgrading. The number of consecutive windows is adaptively determined based on the current deformation rate trend: the microscale index sequence of the five most recent sliding time windows before the current moment is taken, and linear regression is performed on the sequence to obtain the slope k and the goodness of fit R. 2 If the slope k > 0 and the slope significance test p < 0.1 and R < 0.1, then... 2 If the slope p is ≥0.6, it is judged as "continuous acceleration trend", and the number of consecutive windows is 2; if the slope p is ≥0.1 (i.e. not significant) and the coefficient of variation of the micro-scale index CV = standard deviation / |mean| <0.3 (if the absolute value of the mean is less than 0.01, it is considered as stationary fluctuation), it is judged as "stationary fluctuation", and the number of consecutive windows is 5; other cases are judged as "general trend", and the number of consecutive windows is the default value of 3.

[0028] S5 represents the output and handling. This step provides the corresponding output based on the current risk level: At a stable level, maintain the normal data collection frequency and do not issue warnings; at the initial stage, increase the data collection frequency and send a notification to monitoring personnel, prompting them to strengthen manual patrols; at a risk level, issue an evacuation warning and trigger corresponding handling actions according to the plan, such as activating audible and visual alarms, linking gates or road closures, and pushing warning work orders to the emergency platform. The handling content for each level can be configured according to the engineering plan. At this point, S1 to S5 of the current sliding time window have been completed. The monitoring terminal waits for the next time window to slide before starting execution from S1 again.

[0029] Based on the same inventive concept, this invention also provides a multi-scale slope slip index coupling system that integrates macroscopic cracks and microscopic displacements, the module structure of which is as follows: Figure 2 As shown. It should be understood that the following division of each unit is a corresponding division made to implement the functions of the above method. This division is illustrative and is only a logical functional division. In actual implementation, there may be other division methods, and each unit can be implemented by corresponding hardware or software functional modules.

[0030] like Figure 2 As shown, the system includes: a data acquisition unit 301, used to collect the macroscopic crack opening and microscopic displacement of the same monitoring object on the slope and align them according to a unified time reference to obtain macroscopic crack sequences and microscopic displacement sequences; a dual-scale index unit 302, used to calculate the crack propagation rate and displacement rate respectively within the sliding time window and normalize them relative to the baseline fluctuation band to obtain macroscopic scale indices and microscopic scale indices; a cross-scale coupling unit 303, used to calculate the cross-scale consistency coefficient and use it as a gating factor to modulate the microscopic scale indices to obtain cross-scale coupled slip indexes; a classification judgment unit 304, used to compare the cross-scale coupled slip indexes with an adaptively calibrated classification threshold to determine the slip level; and an early warning output unit 305, used to output early warning information or trigger handling actions according to the slip level. In one implementation, the cross-scale coupling unit 303 is also used to set the cross-scale consistency coefficient to 0 and output conservatively when a certain sensor is abnormal (according to the aforementioned obvious fault judgment criteria). The above units are used to perform the corresponding steps of the aforementioned method, and their specific implementation methods and beneficial effects are the same as those of the aforementioned method, and will not be repeated here.

[0031] The present invention also provides an electronic device 40, the structure of which is as follows: Figure 10 As shown, the device includes a processor 41 and a memory 42. The memory 42 stores a computer program, and the processor 41 executes the computer program to implement the steps of the aforementioned method. The electronic device 40 also includes an input device 43 and an output device 44. The input device 43 is used to access monitoring data from a crack gauge, extensometer, displacement gauge, or global navigation satellite system receiver. The output device 44 is used to issue early warning information. The present invention also provides a computer-readable storage medium storing a computer program thereon. When the computer program is executed by a processor, it implements the steps of the aforementioned method. The computer-readable storage medium includes, but is not limited to, read-only memory, random access memory, magnetic disk, or optical disk, and other media capable of storing program code.

[0032] The following is an end-to-end implementation method, using a specific slope monitoring section to illustrate the complete operation of the invention. In this implementation, the evolution of the cross-scale coupled slip index and the cross-scale consistency coefficient during the slope approaching slippage process is as follows: Figure 8As shown in the figure. This implementation method is a sedimentary slope near the dam bank of a hydropower project, whose deformation mode is determined to be traction-type. A crack gauge was installed across the same controlling crack on the slope to collect macroscopic crack opening, and a global navigation satellite system receiver was installed on the same slope monitoring pier to collect microscopic displacement. The parameters in this implementation method are set as follows: sampling interval is 1 hour, sliding time window length is 24 hours (shifted forward every hour), preset backtracking period is 30 days, preset normalization multiple is 4, and the preset consistency zone is set to the P20 and P80 quantiles according to the traction-type distribution, with a calculated lower bound of 0.4 and an upper bound of 2.5. The confirmation threshold is calibrated to 0.6 based on the P80 quantile of the cross-scale consistency coefficient distribution during the stationary period. First, based on the data from the slope's stable period of approximately 30 days, the mean value of the cross-scale coupled slippage index was obtained as 0.3 and the standard deviation as 0.4. Based on this, the first-level threshold was adaptively calibrated to be the mean plus 3 times the standard deviation, i.e., 1.5, and the second-level threshold was the mean plus 8 times the standard deviation, i.e., 3.5.

[0033] The first period was characterized by a heavy rainfall event. Rainfall infiltration caused a general increase in slope displacement readings. Within this window, the normalized microscale index of displacement rate rose to 3.6, while the crack opening measured by the crack gauge remained almost unchanged, with a normalized macroscale index of only 0.15. At this point, the increment ratio between the two scales was approximately 0.04, far below the lower limit of the concordance zone (0.4), indicating that the concordance zone consistency was close to 0. Although the synchronous increase was not zero because both were slightly greater than 0, it was very small due to the extremely disproportionate increment ratio. The calculated cross-scale consistency coefficient was approximately 0.003, which was less than the confirmation threshold of 0.6. Therefore, the microscale index was suppressed, and the cross-scale coupled slippage index was only approximately 0.01, less than the first-level threshold of 1.5, indicating a stable level, and no warning was issued.

[0034] To further illustrate the effects of the present invention, a comparative example 1 is set up to compare with the first time period described above. This comparative example is identical to the first time period except that the cross-scale consistency gating is replaced with the existing equal-weighted superposition. During this period of heavy rainfall, the equal-weighted superposition directly multiplies the macro-scale index 0.15 and the micro-scale index 3.6 by 0.5 and adds them together to obtain a comprehensive index of approximately 1.88. This exceeds the first classification threshold of 1.5, thus falsely reporting a rainfall disturbance that should not have been warned as an initial movement. This is because the equal-weighted superposition lacks cross-scale discrimination to confirm whether the micro-displacement jump is confirmed by the simultaneous expansion of macro-cracks, and therefore cannot distinguish between isolated micro-displacement jumps caused by rainfall and true precursors to landslides, resulting in the fusion value being directly inflated by noise. In contrast, the present invention, by simultaneously confirming macro-cracks, identifies the isolated jump as unsupported, thereby suppressing it to 0.01 and classifying it as a stable level, thus avoiding false alarms. Therefore, it can be seen that the improvement of this invention in that the comparative example does not produce false alarms during this period is directly due to the key constraint of cross-scale consistency gating, rather than the accidental value of the parameter.

[0035] The second period is several weeks later when the slope enters the initial movement state. At this time, the crack gauge and the global navigation satellite system receiver simultaneously measure continuous synchronous deformation. Within this window, the macro-scale index rises to 1.1 and the micro-scale index rises to 1.4. Both are greater than 0 and show an upward trend within the window according to the aforementioned upward trend judgment method. The ratio of the smaller to the larger synchronous increase is approximately 0.79, and the increment ratio is approximately 0.79, falling within the consistency zone of 0.4 to 2.5. The consistency zone conformity is 1, and the cross-scale consistency coefficient is calculated to be approximately 0.79, which is not less than the confirmation threshold of 0.6. Therefore, the micro-scale index is confirmed and enhanced, and the enhancement coefficient is taken as the sum of 1 and the consistency coefficient, i.e., 1.79. The cross-scale coupled slip threshold is approximately 2.50, which is not less than the first-level threshold of 1.5 and less than the second-level threshold of 3.5. It is judged to be in the initial movement stage, and the system increases the acquisition frequency and sends attention information to the monitoring personnel.

[0036] The third time period is when the slope approaches the imminent landslide. Within this window, the macro-scale index rises to 2.6 and the micro-scale index rises to 3.0, both rising synchronously and in the same direction, with a synchronous rise rate of approximately 0.87. The increment ratio of approximately 0.87 remains within the consistency zone, and the consistency zone conformity is 1. The calculated cross-scale consistency coefficient is approximately 0.87, triggering further confirmation enhancement, with an enhancement coefficient of approximately 1.87. The cross-scale coupled landslide index is approximately 5.60, not less than the second-level threshold of 3.5, thus determining it to be at the imminent landslide level. The system issues an evacuation warning and triggers audible and visual alarms and road closures. Comparing the three time periods, it can be seen that this invention suppresses isolated micro-scale jumps to 0.01 during rainfall and noise periods to avoid false alarms. During the initial movement and imminent landslide periods, it uses the synchronous confirmation of macro-cracks to enhance the confirmed micro-deformations and provide early-level warnings, thus balancing early and accurate warnings under the same window-by-window calculation. It should be noted that the above values ​​are illustrative examples of this implementation method. In actual engineering, the corresponding thresholds should be adaptively calibrated based on the monitoring data of this slope.

[0037] To illustrate the effectiveness of this invention, historical monitoring data from a certain slope is used to compare this invention with existing methods. The following data are illustrative examples. Figure 7 As shown, during periods when isolated micro-displacement jumps are caused by environmental disturbances such as rainfall, the false alarm rate of the method based solely on the micro-displacement rate threshold is approximately 35%, while the false alarm rate of the method using equal-weighted superposition of multi-source indicators is approximately 22%. However, this invention reduces the false alarm rate to approximately 4% during these periods by simultaneously confirming macroscopic cracks. This is because this invention suppresses isolated micro-displacement jumps that are not confirmed by macroscopic cracks, thus preventing them from being triggered by environmental noise. Figure 8As shown, as the slope progresses from the stable phase through the initial movement phase and approaches the slippage stage, the cross-scale coupled slippage index increases from approximately 0.3 to approximately 2.50 and then to approximately 5.60, while the cross-scale consistency coefficient increases from approximately 0.05 to approximately 0.79 and then to approximately 0.87, with both rising synchronously. Based on this, the present invention provides graded early warnings during the initial movement phase, which, compared to the conventional practice of waiting for a significant acceleration of macroscopic cracks before issuing an alarm, allows for earlier warnings without sacrificing reliability. All the above values ​​are illustrative examples and do not constitute a limitation on the scope of protection of this invention.

[0038] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for coupling macroscopic cracks and microscopic displacements into a multi-scale slope slip index, characterized in that, Includes the following steps: S1, during the first run, obtain the macroscopic crack opening amount and microscopic displacement amount during the stable period of this slope, and use them to determine the verification threshold and the classification threshold. Subsequently, the macroscopic crack opening and microscopic displacement of the same monitoring object on the slope were collected and aligned according to a unified time reference to obtain the macroscopic crack sequence and microscopic displacement sequence. S2, calculate the crack propagation rate of the macro crack sequence and the displacement rate of the micro displacement sequence within the sliding time window, and normalize the crack propagation rate and the displacement rate respectively with the baseline fluctuation band of the crack propagation rate and the displacement rate within the preset backtracking period as the benchmark to obtain dimensionless macro-scale index and micro-scale index. S3, calculate the cross-scale consistency coefficient between the macro-scale index and the micro-scale index within the sliding time window; the cross-scale consistency coefficient is positively correlated with the degree to which the macro-scale index and the micro-scale index rise synchronously in the same direction, and is also positively correlated with the degree to which the first increment ratio falls into the preset consistency band; the first increment ratio is the ratio of the macro-scale index to the micro-scale index, and the preset consistency band is a pre-set second increment ratio interval; The microscale index is modulated using the cross-scale consistency coefficient as a gating factor. If the cross-scale consistency coefficient is not less than the confirmation threshold, the microscale index is enhanced for confirmation; otherwise, it is suppressed, thus obtaining the cross-scale coupled slippage index. S4, compare the cross-scale coupled slippage index with the classification threshold, and determine the slippage level after time series consistency verification; S5, output early warning information or trigger corresponding handling actions according to the aforementioned slip level.

2. The method according to claim 1, characterized in that, The macroscopic crack opening is collected by a crack gauge or expansion gauge installed at the monitoring object, and the microscopic displacement is collected by a displacement gauge or a global navigation satellite system receiver installed at the same monitoring object. The alignment according to a unified time reference includes resampling the macroscopic crack opening amount and the microscopic displacement amount to a common sampling interval, and filling the missing measurement points with the nearest neighbor time.

3. The method according to claim 1, characterized in that, The baseline fluctuation band is determined by the median and median absolute deviation of the crack propagation rate or the displacement rate during the preset backtracking period. The normalization is as follows: for the crack propagation rate, subtract the median of the crack propagation rate within the preset backtracking period, and then divide by the product of a preset multiple and the absolute deviation of the median of the crack propagation rate to obtain the macroscopic scale index; for the displacement rate, normalize in the same way to obtain the microscopic scale index; the preset multiple ranges from not less than 3 to not greater than 6.

4. The method according to claim 1, characterized in that, The cross-scale consistency coefficient is the product of the degree to which the macro-scale index and the micro-scale index rise synchronously and in the same direction and the degree to which the first increment falls into the preset consistency band. The degree to which the macro-scale indicator and the micro-scale indicator rise synchronously and in the same direction is determined using the following method: When both the macro-scale index and the micro-scale index are greater than 0, and both show an upward trend within the sliding time window, the ratio of the smaller of the two to the larger of the two is taken; otherwise, 0 is taken. The term "all showing an upward trend" means that the end value of the sequence is greater than the start value, and the linear slope of the least squares fit to the sequence is positive and passes the significance test; the degree of conformity of the conformity band is 1 when the increment ratio falls within the preset conformity band, and monotonically decreases to 0 as the increment ratio deviates from the preset conformity band when it falls outside the preset conformity band.

5. The method according to claim 4, characterized in that, In the step of modulating the microscale index using the cross-scale consistency coefficient as a gating factor, and enhancing the microscale index if the cross-scale consistency coefficient is not less than the confirmation threshold, and suppressing it otherwise, to obtain the cross-scale coupled slippage index: When the cross-scale consistency coefficient is not less than the confirmation threshold, the microscale index is multiplied by the enhancement coefficient to obtain the cross-scale coupled slip index, where the enhancement coefficient α = 1 + γ, and γ is the cross-scale consistency coefficient. When the cross-scale consistency coefficient is less than the confirmation threshold, the microscale index is multiplied by the inhibition coefficient to obtain the cross-scale coupled slip index, where the inhibition coefficient β=γ, which is equal to the cross-scale consistency coefficient.

6. The method according to claim 4, characterized in that, The upper and lower boundaries of the preset consistency zone are determined by the deformation ratio between macroscopic crack propagation and microscopic displacement corresponding to the deformation mode of this slope. The deformation mode includes traction and shoving. The upper and lower boundaries are uniformly determined by the quantiles of the increment ratios of each sliding time window in which both macroscopic and microscopic scale indices are greater than 0 during the stable period of this slope. The quantile interval for traction slopes is wider than that for shoving slopes. The confirmation threshold is determined by the quantiles of the cross-scale consistency coefficient distribution during the stable period of this slope, and the value range is not less than 0.5 and not greater than 0.

8.

7. The method according to claim 1, characterized in that, The grading thresholds include a first grading threshold and a second grading threshold, which are adaptively calibrated by the mean and standard deviation of the cross-scale coupled slippage index during the stable period of this slope according to a preset multiple, and the second grading threshold is greater than the first grading threshold. The determination of the slip risk level is as follows: when the cross-scale coupled slip risk index is less than the first classification threshold, it is determined to be a stable level; when it is not less than the first classification threshold and less than the second classification threshold, it is determined to be an initial movement level; when it is not less than the second classification threshold, it is determined to be a slip risk level. The upgrade requirement for the slip risk level is that the level is consistent within the continuous sliding time window. The number of windows in the continuous sliding time window is adaptively determined according to the current deformation rate trend. The S5 outputs warning information or triggers response actions according to the level of skid approach, including encrypting and collecting information of concern at the initial stage, and issuing evacuation warnings and triggering response actions at the skid approach level.

8. A multi-scale slope slip index coupling system integrating macroscopic cracks and microscopic displacement, characterized in that, include: The data acquisition unit is used to collect the macroscopic crack opening amount and microscopic displacement amount of the same monitoring object on the slope, and align them according to a unified time reference to obtain the macroscopic crack sequence and microscopic displacement sequence. The dual-scale index unit is used to calculate the crack propagation rate of the macro crack sequence and the displacement rate of the micro displacement sequence within a sliding time window, and normalize them respectively based on the baseline fluctuation band within a preset backtracking period to obtain macro-scale index and micro-scale index. A cross-scale coupling unit is used to calculate the cross-scale consistency coefficient between the macro-scale index and the micro-scale index, and to use the cross-scale consistency coefficient as a gating factor to modulate the micro-scale index to obtain a cross-scale coupled slip index. The classification and determination unit is used to compare the cross-scale coupled landslide risk index with the classification threshold adaptively calibrated by the data of the stable period of this slope to determine the landslide risk level; The early warning output unit is used to output early warning information or trigger corresponding handling actions according to the aforementioned level of slippage.

9. A computer-readable storage medium, characterized in that, The storage medium stores a computer program that, when executed by a processor, implements the steps of the method according to any one of claims 1 to 7.

10. An electronic device, characterized in that, It includes a memory and a processor, wherein the memory stores a computer program that, when executed by the processor, implements the steps of the method according to any one of claims 1 to 7.

Images