A data processing method for deformation monitoring of rock-soil mass

Abstract

The present application belongs to the technical field of data processing, and particularly relates to a data processing method for rock-soil mass deformation monitoring. The method first acquires multi-monitoring point displacement time series data to construct a space-time displacement field; then extracts deformation field topological features through field analysis, identifies deformation energy source and sink points and streamline networks, and calculates monitoring point topological role weights; then screens instability-related feature frequency bands through multi-scale wavelet transform, constructs a phase coupling matrix, divides sub-regions, and calculates coupling indexes; finally, fuses the topological role weights and the phase coupling indexes to construct a deformation state entropy model, and triggers graded early warning according to the entropy value change rate. The present application breaks through the limitations of traditional single-point analysis, accurately captures the deformation propagation law, improves the instability precursor recognition ability and early warning reliability, and adapts to the multi-dimensional deformation monitoring needs of complex geology.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of data processing, and particularly relates to a data processing method for deformation monitoring of a rock-soil mass. BACKGROUND

[0002] Deformation monitoring of a rock-soil mass is a core link of safety management and control of a rock-soil engineering such as tunnel construction, and accurate deformation data processing is a key to identifying precursors of instability and avoiding collapse disasters. The traditional monitoring method relies on changes in single-point displacement curves for early warning, and there is an obvious bottleneck in dealing with complex geological conditions and multi-dimensional deformation propagation mechanisms. With the development of sensing technology, a monitoring system can synchronously collect three-dimensional displacement time series data of multiple monitoring points, but the spatiotemporal correlation of massive data is complex, and how to mine deformation evolution rules from the data becomes a technical difficulty, which makes it difficult to ensure reliable early warning. SUMMARY

[0003] The application proposes a data processing method for deformation monitoring of a rock-soil mass in view of the technical problems in the above background technology.

[0004] In order to achieve the above purpose, the technical scheme adopted by the application comprises the following steps:

[0005] Obtain displacement time series data of multiple monitoring points in a monitoring area, and construct a spatiotemporal displacement field;

[0006] Perform field analysis on the spatiotemporal displacement field, extract topological feature structures of the deformation field, identify source point areas, sink point areas of deformation energy, and deformation streamline networks connecting the two;

[0007] Based on the spatial distribution characteristics of the topological feature structures, calculate the topological role weights of the monitoring points in the deformation propagation path, and the topological role weights represent the influence degree of the monitoring points on the overall deformation field evolution;

[0008] Perform multi-scale wavelet transform on the displacement time series data of each monitoring point, calculate the time-frequency coupling index of phase consistency and deformation energy, filter feature frequency bands related to rock-soil mass instability, construct a phase coupling matrix based on the feature frequency bands, divide the monitoring area into sub-areas by applying a community detection algorithm, and calculate the phase coupling index of each sub-area;

[0009] Fuse the topological role weights of each monitoring point and the phase coupling index of each sub-area to construct a deformation state entropy model, analyze the entropy value time series evolution trajectory, and trigger a deformation risk early warning mechanism of a corresponding level when the entropy value change rate continuously exceeds a preset hierarchical threshold.

[0010] As a preferred, the implementation of obtaining displacement time series data of multiple monitoring points in a monitoring area and constructing a spatiotemporal displacement field comprises:

[0011] N displacement detection points with fixed spatial positions are arranged in the monitoring area;

[0012] synchronous displacement data collection is performed on each monitoring point at a preset sampling frequency, to obtain three-dimensional displacement components of each monitoring point i in a time sequence , wherein, are respectively components of displacement of the i th monitoring point in x, y and z directions at time t;

[0013] coordinate system unification and reference surface normalization are performed on the displacement data, to structurally organize displacement data of all monitoring points at each time according to spatial positions and time stamps, to form a four-dimensional data set D ; wherein, is a spatial coordinate of the i th monitoring point;

[0014] based on the four-dimensional data set D, displacement field reconstruction is performed between monitoring points through a space-time interpolation algorithm, to generate a continuous space-time displacement field function , wherein, is a space-time weight coefficient, is an arbitrary spatial point coordinate in the monitoring area, is an adaptive spatial scale parameter; so that for any spatial position p and any time t, a corresponding displacement estimate can be obtained , so as to complete construction of the space-time displacement field.

[0015] As a preferred, field analysis is performed on the space-time displacement field, to extract topological feature structures of the deformation field, to identify source point areas, sink point areas and deformation streamline network connecting the two, which includes:

[0016] a displacement gradient vector field is constructed , wherein, is a gradient operator, a deformation energy density function is calculated based on the displacement gradient vector field ; key topological structures in the deformation field are identified, including source point areas, sink point areas and saddle point areas; the source point area is a spatial area satisfying and in a neighborhood of 0, wherein, is a source point determination threshold, respectively represent divergence and curl operators; the sink point area is a spatial area satisfying and in a neighborhood of 0; the saddle point area is a spatial area satisfying in a neighborhood of 0 and ;

[0017] ​Starting from the source region, the deformable streamlines are traced along the direction of the displacement gradient vector, where the streamline tracing step size is... With deformation energy density Adaptive adjustment: ,in, These are the minimum and maximum step sizes, respectively. For adaptive coefficients, Energy threshold;

[0018] The streamline termination condition is determined when the deformable streamline reaches the sink region, or when the deformation energy density at the location of the deformable streamline is determined. When the energy level is less than the preset minimum energy threshold, the tracking of the deformed streamline stops, the construction of a single deformed streamline is completed, and a deformed streamline network is formed.

[0019] Preferably, based on the spatial distribution characteristics of the topological feature structure, the topological role weight of each monitoring point in the deformation propagation path is calculated. The topological role weight characterizes the degree of influence of the monitoring point on the overall deformation field evolution, including:

[0020] Based on deformable streamline networks, energy transfer efficiency is selected. The highest front The streamlines serve as the primary deformation propagation path, among which, Let L be the energy transfer efficiency of the streamline, where L is the length of the streamline. For integration variables;

[0021] If monitoring point i is located on any of the main deformation propagation paths, then calculate its Euclidean distance to the nearest source region. And calculate the initial topology role weights. ,in To avoid small constants that divide by zero; if monitoring point i is not located on any major deformation propagation path, the topological role weight is 0;

[0022] The initial topological role weights of each monitoring point are normalized to obtain the topological role weights.

[0023] As a preferred approach, multi-scale wavelet transform is performed on the displacement time-series data of each monitoring point to calculate the time-frequency coupling index of phase consistency and deformation energy. Characteristic frequency bands related to soil instability are selected, a phase coupling matrix is ​​constructed based on these characteristic frequency bands, and a community detection algorithm is applied to divide the monitoring area into sub-regions. The specific implementation of calculating the phase coupling index for each sub-region includes:

[0024] The displacement time series data of the i-th monitoring point are obtained using the Morlet wavelet basis. Perform multi-scale continuous wavelet transform to obtain the time-frequency domain representation. where s is a scale parameter, which is automatically optimized by the maximal overlap discrete wavelet transform, satisfying is less than a preset convergence threshold;

[0025] The phase consistency index PC(s, t) and the deformation energy index E(s, t) at scale s are calculated: , where J is an imaginary unit, is the phase angle of the i th monitoring point at scale s and time t, N is the total number of monitoring points in the monitoring area, is the topological role weight of the i th monitoring point, is a modulo operation;

[0026] The time-frequency coupling index is constructed and the characteristic frequency band is screened, and the cumulative coupling strength of each scale s is calculated where is the start time, and when is greater than a set threshold and lasts for more than a preset time, the frequency band corresponding to the scale s is determined as a characteristic frequency band related to the instability of the rock-soil mass;

[0027] Based on the identified characteristic frequency band, the phase time series of each monitoring point in the frequency band is extracted The weighted phase coupling matrix C is constructed, where the element in the matrix is The implementation of the element in the matrix is to first calculate the phase locking value of the monitoring points i and j within the time window, i.e., to calculate the average value of the complex exponential function for all time points t, introduce a spatial attenuation factor, calculate the spatial Euclidean distance of the monitoring points i and j , and apply an exponential decay function for weighting, where is a spatial attenuation coefficient, representing the spatial influence range of the rock-soil mass deformation, and finally the phase locking value is multiplied by the spatial attenuation factor to obtain , representing the phase synchronization strength of the monitoring points i and j considering the influence of spatial distance;

[0028] The phase coupling matrix C is divided into sub-regions by a community detection algorithm, and spatially continuous sub-regions are obtained, and the phase coupling index of the k th sub-region is calculated: where represents the k th sub-region, is a modulo operation.

[0029] As preferred, the topological role weight of each monitoring point and the phase coupling index of each sub-region are fused to construct a deformation state entropy model, and the time sequence evolution trajectory of the entropy value is analyzed, and when the change rate of the entropy value continuously exceeds the preset grading threshold, the implementation of the corresponding level of the deformation risk early warning mechanism includes:

[0030] For the kth region in the sub-region, the topological role weight of each monitoring point in the region is normalized to obtain the probability distribution , and the local deformation entropy of the sub-region k is calculated , wherein is the phase coupling index of the sub-region k;

[0031] According to the local deformation entropy of the sub-region k , the global deformation state entropy value is obtained ; the entropy value change rate of adjacent time points is calculated , and if the entropy value change rate of three consecutive time points is less than 0 and lower than the set threshold, the early warning is determined.

[0032] Compared with the prior art, the advantages and positive effects of the present application are that the traditional early warning limitation depending on single point displacement curve is broken through, the continuous reconstruction of the deformation state is realized by constructing the space-time displacement field, the spatial correlation and propagation characteristics of the rock-soil mass deformation are accurately captured, and the one-sided misjudgment caused by local data is avoided; the topological characteristics of the deformation field are deeply mined by means of field theory analysis, the source and sink points and the streamline network of the deformation energy are clearly determined, the deformation evolution mechanism is clearly revealed, and the pain point that the traditional method is difficult to grasp the overall deformation rule is solved; the topological role weight and the multi-scale time-frequency coupling analysis are innovatively fused, the instability related characteristic frequency band is effectively selected, the noise interference is greatly suppressed, and the key signal recognition degree is improved; by constructing the deformation state entropy model, combining the phase coupling characteristics of the sub-region, and based on the entropy value change rate, the graded early warning is realized, the timeliness and reliability of the early warning are significantly enhanced, and the multi-dimensional deformation monitoring demand under complex geological conditions can be better adapted. BRIEF DESCRIPTION OF DRAWINGS

[0033] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiment description. Obviously, the drawings in the following description are some embodiments of the present application, and other drawings can also be obtained by those skilled in the art without creative labor.

[0034] Figure 1 It is a structural flowchart of a data processing method for rock-soil mass deformation monitoring. DETAILED DESCRIPTION

[0035] To better understand the above-mentioned objectives, features, and advantages of the present invention, the present invention will be further described below in conjunction with the accompanying drawings and embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in these embodiments can be combined with each other.

[0036] Numerous specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways than those described herein, and therefore the invention is not limited to the specific embodiments disclosed in the following specification.

[0037] In geotechnical engineering projects such as tunnel excavation, the internal structure of soil and rock masses is complex, and their deformation processes often exhibit strong nonlinearity, heterogeneity, and abrupt changes. Traditional monitoring methods rely on changes in single-point displacement curves for early warning, which has significant limitations when dealing with complex geological conditions and multidimensional deformation propagation mechanisms. To identify instability signs earlier and provide reliable early warnings, it is necessary to integrate spatiotemporal data from multiple monitoring points, topological propagation characteristics, and energy evolution mechanisms to construct a more comprehensive deformation cognition model. Therefore, a data processing method for monitoring soil and rock deformation is proposed, specifically implemented as follows: Figure 1 As shown.

[0038] Firstly, to reveal the energy transfer path and key control areas of soil and rock deformation, and to address the issue of existing methods focusing only on single-point displacement values ​​while neglecting the overall structural characteristics of the deformation field, this method acquires time-series displacement data from multiple monitoring points within the monitoring area to construct a spatiotemporal displacement field. This is achieved by deploying N spatially fixed displacement detection points within the monitoring area; synchronously acquiring displacement data from each monitoring point at a preset sampling frequency; and obtaining the three-dimensional displacement components of each monitoring point i in the time series. ,in, The i-th monitoring point is located at The displacement components in the x, y, and z directions at each moment are analyzed. The displacement data are normalized using coordinate system one and a reference plane. The displacement data of all monitoring points at each moment are then structured according to spatial location and timestamp to form a four-dimensional dataset. ;in, Let be the spatial coordinates of the i-th monitoring point; based on the four-dimensional dataset D, a spatiotemporal interpolation algorithm is used to reconstruct the displacement field between the monitoring points, generating a continuous spatiotemporal displacement field function. ,in, For spatiotemporal weighting coefficients, To monitor the coordinates of any spatial point within the area, The parameters are adaptive spatial scale parameters, ensuring that a displacement estimate can be obtained for any spatial location p and any time t. This completes the construction of the spatiotemporal displacement field.

[0039] Specifically, N monitoring points with fixed spatial positions are deployed within the target rock-soil mass region, and the three-dimensional displacement components of each monitoring point at each time are synchronously collected at a preset sampling frequency, thereby obtaining discrete multi-point time-series data reflecting the real deformation process. Subsequently, coordinate system standardization and datum surface normalization processing are performed on all monitoring point data to eliminate interference such as equipment differences, sensing errors, and inconsistent spatial references, so that the data have comparability and uniformity. The arranged monitoring point displacement data are organized into a four-dimensional data set according to spatial positions and time stamps, laying a foundation for subsequent field function construction. Based on the four-dimensional data set, a spatio-temporal interpolation algorithm containing a spatial self-adaptive scale parameter is introduced to reconstruct a continuous spatio-temporal displacement field between the monitoring points, so that the displacement estimation value of any spatial position p and any time t can be obtained. The spatio-temporal displacement field constructed by the above method not only realizes the reconstruction of the deformation state from discrete monitoring points to continuous space, but also provides a data basis for subsequent deformation gradient analysis, energy density calculation, and topological feature extraction. Its advantages are: it can reveal the spatial correlation of the internal deformation of the rock-soil mass from a holistic perspective, restore the continuity and propagation of displacement changes, avoid the local one-sidedness caused by the limited number of monitoring points, and thus significantly improve the accuracy and integrity of the deformation mechanism analysis, providing a reliable basis for identifying the source, sink and propagation path of deformation energy.

[0040] Then considering that the traditional method cannot identify the source, sink and propagation path of deformation energy, leading to misjudgment of the deformation evolution mechanism. Field analysis is performed on the spatio-temporal displacement field, and the topological feature structure of the deformation field is extracted to identify the source area, sink area and deformation streamline network connecting the two. The implementation includes constructing a displacement gradient vector field wherein, is a gradient operator, calculating a deformation energy density function based on the displacement gradient vector field; identifying key topological structures in the deformation field, including source areas, sink areas and saddle point areas; the source area satisfies and is a spatial area within the left and right neighborhoods of the set value 0, wherein, is a source point determination threshold, respectively represent the divergence and curl operators; the sink area satisfies and is a spatial area within the left and right neighborhoods of the set value 0; the saddle point area satisfies is a spatial area within the left and right neighborhoods of the set value 0 and ; starting from the source area, the deformation streamline is traced along the displacement gradient vector direction, wherein the streamline tracing step is adaptively adjusted with the deformation energy density : wherein, These are the minimum and maximum step sizes, respectively. For adaptive coefficients, The energy threshold is used to determine the streamline termination condition, which is when the deformed streamline reaches the sink region, or the deformation energy density at the location of the deformed streamline is determined. When the energy level is less than the preset minimum energy threshold, the tracking of the deformed streamline stops, the construction of a single deformed streamline is completed, and a deformed streamline network is formed.

[0041] Specifically, the displacement gradient vector field is calculated based on the spatiotemporal displacement field function. Based on this, the deformation energy density distribution is further derived from the gradient field. After obtaining the energy field, key topological structures in the deformation field of the soil and rock mass are identified through a comprehensive judgment of divergence, curl, and energy density. Regions satisfying divergence greater than a set threshold and curl close to zero are defined as energy release source regions; regions satisfying divergence less than the opposite threshold and curl close to zero are defined as energy convergence sink regions; and regions satisfying both divergence and curl close to zero but with abrupt changes in gradient direction are identified as saddle points. These regions are often key locations for deformation path inflection or energy redistribution. To reveal the propagation trajectory of deformation energy in space, streamline tracing is performed along the displacement gradient direction starting from each source region. Streamline tracing employs an adaptive step-size strategy: the step size is reduced when the local energy density is high to enhance path accuracy, and the step size is increased when the energy density is low to improve computational efficiency. The tracing process terminates when the streamline reaches the sink region or the energy density drops to the lowest threshold, thus obtaining a complete deformation streamline. By repeatedly tracing all source point regions, a deformation streamline network covering the entire monitoring area is ultimately constructed. Through the above field theory and topological analysis process, the deformation is transformed from a traditional discrete point change to a continuous field structure analysis.

[0042] Next, to accurately extract time-frequency features directly related to soil instability and address the limitations of existing methods in distinguishing effective signals from noise interference and locating instability precursor frequency bands in frequency domain analysis, the topological role weight of each monitoring point in the deformation propagation path is calculated based on the spatial distribution characteristics of the topological feature structure. This topological role weight characterizes the degree of influence of each monitoring point on the overall deformation field evolution. This is achieved by selecting energy transfer efficiency based on the deformation streamline network. The highest front The streamlines serve as the primary deformation propagation path, among which, Let L be the energy transfer efficiency of the streamline, where L is the length of the streamline. Let i be the integral variable; if monitoring point i is located on any major deformation propagation path, then calculate its Euclidean distance to the nearest source region. And calculate the initial topology role weights. ,in To avoid a small constant except zero; if the monitoring point i is not located on any main deformation propagation path, the topology role weight is 0; the initial topology role weight of each monitoring point is normalized to obtain the topology role weight.

[0043] Specifically, based on the constructed deformation streamline network, all streamlines are sorted in descending order of energy transmission efficiency. Energy transmission efficiency is represented by the integral of energy density on the streamline length, which can represent the continuity and intensity of energy propagation along the path. The top several streamlines with the highest energy transmission efficiency are selected as the main deformation propagation paths, which correspond to the most significant deformation evolution channels in the rock-soil mass and have a decisive influence on the instability process. Subsequently, it is determined whether each monitoring point is located on the above main propagation paths. If a monitoring point falls on any main streamline, the Euclidean distance between the monitoring point and the nearest source point region is further calculated. The shorter the distance, the closer the monitoring point is to the energy starting position, and the more sensitive the deformation signal carried by the monitoring point is to the overall energy evolution. Therefore, the initial topology role weight is constructed in the form of the reciprocal of the distance to avoid introducing a small constant when dividing by zero and to ensure stable and reliable weight calculation. For monitoring points that do not fall on the main streamlines, their role in deformation propagation is relatively weak, so their topology role weight is set to zero. Finally, the initial weights of all monitoring points are normalized to form a topology role weight vector. This vector not only preserves the relative importance of different monitoring points in the propagation structure, but also can be deeply integrated with the subsequent time-frequency feature extraction to achieve structure-signal dual enhancement. The above process changes the importance evaluation of monitoring points from displacement amplitude to structural contribution by introducing deformation streamline structure information, greatly improving the scientificity of signal screening; secondly, the weight can be used to improve the signal-to-noise ratio of key frequency bands in time-frequency analysis, strengthen the real instability precursor features, and suppress random noise.

[0044] To accurately extract the time-frequency features directly related to the instability of rock-soil mass and solve the problem that existing methods cannot distinguish between effective signals and noise interference in frequency domain analysis and cannot locate the instability precursor feature frequency band, for the displacement time series data of each monitoring point, multi-scale wavelet transform is performed, the time-frequency coupling index of phase consistency and deformation energy is calculated, the feature frequency band related to the instability of rock-soil mass is selected, the phase coupling matrix is constructed based on the feature frequency band, the community detection algorithm is applied to divide the monitoring area into sub-areas, and the phase coupling index of each sub-area is calculated, which includes Performing multi-scale continuous wavelet transform to obtain time-frequency domain representation where s is the scale parameter, which is automatically optimized by maximum overlap discrete wavelet transform, satisfying Less than a preset convergence threshold; calculate the phase consistency index PC(s, t) and the deformation energy index E(s, t) under the scale s: , where J is the imaginary unit, is the phase angle of the i th monitoring point at scale s and time t, and N is the total number of monitoring points in the monitoring area, is the topological role weight of the i th monitoring point, is a modulo operation; construct the time-frequency coupling index , and select the characteristic frequency band, calculate the cumulative coupling strength of each scale s where, is the starting time, when is greater than the set threshold and lasts for more than the preset time, it is determined that the frequency band corresponding to the scale s is the characteristic frequency band related to the instability of the rock-soil mass; based on the identified characteristic frequency band, the phase time series of each monitoring point in the frequency band is extracted construct a weighted phase coupling matrix C, where the element in the matrix is The implementation of the element in the matrix is to first calculate the phase locking value of the monitoring points i and j within the time window, that is, to calculate the average value of the complex exponential function for all time points t, introduce a spatial attenuation factor, calculate the spatial Euclidean distance between monitoring points i and j, and apply an exponential attenuation function for weighting, where is the spatial attenuation coefficient, representing the spatial influence range of the rock-soil mass deformation, and finally the product of the phase locking value and the spatial attenuation factor is obtained , representing the phase synchronization strength of monitoring points i and j considering the influence of spatial distance; through the community detection algorithm, the phase coupling matrix C is divided into sub-regions with good spatial continuity, and the phase coupling index of the k th sub-region is calculated: where represents the k th sub-region, is a modulo operation.

[0045] Specifically, the original displacement time series of i monitoring points are subjected to multi-scale continuous wavelet transform using Morlet wavelet basis. Morlet wavelet has good time-frequency localization performance, which can capture short-time burst signals and identify long-term trends. By setting multiple scale parameters s, the deformation behavior in different frequency bands is decomposed. In order to ensure the rationality of scale selection, the maximum overlap discrete wavelet transform (MODWT) is introduced to automatically optimize the scale parameters, so that the energy distribution difference between adjacent scales is less than the preset convergence threshold, thereby obtaining the scale sequence that best represents the deformation characteristics. After wavelet transform, the phase consistency index and deformation energy index at scale s and time t are calculated. The phase consistency is used to identify whether there is a coordinated change behavior between different monitoring points at the same scale; the deformation energy represents the strength of displacement fluctuation at this scale. The phase angle is obtained from the complex form of the wavelet coefficient. To enhance the contribution of important monitoring points, the topological role weight is introduced when calculating the phase consistency index, so that the key control points have higher influence in the coupling analysis. To select the main frequency band related to instability, the time-frequency coupling index is constructed, and the cumulative coupling strength of the phase consistency index and the deformation energy index at each scale is integrated to obtain the cumulative coupling strength of scale s. When this strength continuously exceeds the set threshold and lasts for a preset time length, the frequency band corresponding to this scale is determined as the instability-sensitive characteristic frequency band. This mechanism can automatically filter noise-dominated scales and highlight structural deformation characteristics. Subsequently, based on the identified characteristic frequency band, the phase time series of each monitoring point at this frequency band are extracted to construct the weighted phase coupling matrix C. The calculation process of the matrix elements includes three steps: first, calculate the complex exponential average of the phase difference between monitoring points i and j within the time window, i.e. the phase locking value; second, calculate the spatial Euclidean distance between the two monitoring points; finally, multiply the phase locking value by the spatial attenuation factor. This weighting method can more truly reflect the deformation correlation between spatially close monitoring points and avoid false positives caused by accidental synchronization between distant monitoring points. The constructed coupling matrix C can be regarded as a weighted network, through community detection algorithm, Louvain modularity optimization algorithm can be selected, through iteratively maximizing the network modularity, monitoring points with strong coupling strength are automatically aggregated into the same community; spectral clustering method can also be used, by calculating the eigenvectors of the Laplacian matrix, the clustering structure of the network in high-dimensional space is projected into separable sub-regions. The above algorithm can automatically form multiple spatially continuous, internally highly synchronized regional divisions based on coupling strength, realizing the structured expression of deformation coordination in the monitoring area. The monitoring area is divided into several spatially continuous, internally highly synchronized sub-regions. For each sub-region k, the phase coupling index is further calculated by taking the weighted average of the coupling strength between all monitoring points in the sub-region, to quantify the consistency of the deformation behavior in the region.

[0046] Finally, a deformation state entropy model is constructed by integrating the topological role weights of each monitoring point and the phase coupling index of each sub-region. The temporal evolution trajectory of the entropy value is analyzed, and a corresponding level of deformation risk warning mechanism is triggered when the rate of change of the entropy value continuously exceeds a preset grading threshold. Its implementation includes, for For the k-th region among the sub-regions, normalize the topological role weights of each monitoring point within it to obtain the probability distribution. Calculate the local deformation entropy of subregion k. ,in Let be the phase coupling index of subregion k; based on the local deformation entropy of subregion k. Obtain the global deformation state entropy value The core of global deformation state entropy is to measure the disorder and uncertainty of the system state. When entropy is high (disordered), deformation at monitoring points is randomly dispersed and phase-dependent, the deformation field lacks a unified development trend, and energy does not concentrate on a particular path. When entropy is low (ordered), deformation at monitoring points is highly synchronized, energy concentrates along a specific path, and the deformation field forms a unified evolution direction, corresponding to precursors of instability. Therefore, establishing an entropy time series and calculating the rate of change of entropy at adjacent time points is crucial. If three consecutive time points Warnings are issued when the value is less than 0 and below the set threshold.

[0047] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments for application in other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A data processing method for monitoring deformation of soil and rock masses, characterized in that, Includes the following steps: Obtain displacement time-series data from multiple monitoring points in the monitoring area to construct a spatiotemporal displacement field; Field theory analysis is performed on the spatiotemporal displacement field to extract the topological features of the deformation field and identify the source and sink regions of deformation energy, as well as the deformation streamline network connecting the two. Based on the spatial distribution characteristics of the topological feature structure, the topological role weight of each monitoring point in the deformation propagation path is calculated. The topological role weight represents the degree of influence of the monitoring point on the overall deformation field evolution. For the displacement time series data of each monitoring point, multi-scale wavelet transform is performed to calculate the time-frequency coupling index of phase consistency and deformation energy, select characteristic frequency bands related to soil instability, construct a phase coupling matrix based on the characteristic frequency bands, apply the community detection algorithm to divide the monitoring area into sub-regions, and calculate the phase coupling index of each sub-region. A deformation state entropy model is constructed by integrating the topological role weights of each monitoring point and the phase coupling index of each sub-region. The temporal evolution trajectory of the entropy value is analyzed. When the rate of change of the entropy value continuously exceeds the preset classification threshold, the deformation risk warning mechanism of the corresponding level is triggered. The implementation of acquiring displacement time-series data of multiple monitoring points in the monitoring area and constructing a spatiotemporal displacement field includes: N displacement detection points with fixed spatial locations are set up within the monitoring area; Simultaneous displacement data is collected at each monitoring point using a preset sampling frequency to obtain the three-dimensional displacement components of each monitoring point i in the time series. ,in, The i-th monitoring point is located at The components of displacement in the x, y, and z directions at time t; The displacement data is normalized to coordinate system 1 and reference surface, and the displacement data of all monitoring points at each time point are structured according to spatial location and timestamp to form a four-dimensional dataset. ;in, Let i be the spatial coordinates of the i-th monitoring point; Based on the four-dimensional dataset D, a spatiotemporal interpolation algorithm is used to reconstruct the displacement field between monitoring points, generating a continuous spatiotemporal displacement field function. ,in, For spatiotemporal weighting coefficients, To monitor the coordinates of any spatial point within the area, The parameters are adaptive spatial scale parameters, ensuring that a displacement estimate can be obtained for any spatial location p and any time t. Thus, the spatiotemporal displacement field is constructed.

2. The data processing method for monitoring deformation of soil and rock mass according to claim 1, characterized in that, The implementation of field theory analysis of the spatiotemporal displacement field, extraction of the topological features of the deformation field, and identification of the source and sink regions of deformation energy, as well as the deformation streamline network connecting the two, includes: Constructing the displacement gradient vector field ,in, As a gradient operator, the deformation energy density is calculated based on the displacement gradient vector field. ; The key topological structures in the deformable field are identified, including source point regions, sink point regions, and saddle point regions; the source point region is defined as satisfying... and Within the spatial regions defined as the left and right neighborhoods of 0, where, The threshold for determining the source point. These represent the divergence and curl operators, respectively; the sink region is defined as satisfying... and The spatial region within the left and right neighborhoods of 0; the saddle point region is defined as satisfying... Within the defined left and right neighborhoods of 0 and Spatial region; Starting from the source region, the deformable streamlines are traced along the direction of the displacement gradient vector, where the streamline tracing step size is... With deformation energy density Adaptive adjustment: ,in, These are the minimum and maximum step sizes, respectively. For adaptive coefficients, Energy threshold; The streamline termination condition is determined when the deformable streamline reaches the sink region, or when the deformation energy density at the location of the deformable streamline is determined. When the energy level is less than the preset minimum energy threshold, the tracking of the deformed streamline stops, the construction of a single deformed streamline is completed, and a deformed streamline network is formed.

3. The data processing method for monitoring deformation of soil and rock mass according to claim 2, characterized in that, Based on the spatial distribution characteristics of the aforementioned topological features, the topological role weight of each monitoring point in the deformation propagation path is calculated. This topological role weight characterizes the degree of influence of the monitoring point on the overall deformation field evolution, including: Based on deformable streamline networks, energy transfer efficiency is selected. The highest front The streamlines serve as the primary deformation propagation path, among which, Let L be the energy transfer efficiency of the streamline, where L is the length of the streamline. For integration variables; If monitoring point i is located on any of the main deformation propagation paths, then calculate its Euclidean distance to the nearest source region. And calculate the initial topology role weights. ,in To avoid small constants that divide by zero; if monitoring point i is not located on any major deformation propagation path, the topological role weight is 0; The initial topological role weights of each monitoring point are normalized to obtain the topological role weights.

4. The data processing method for monitoring deformation of soil and rock mass according to claim 3, characterized in that, For the displacement time-series data of each monitoring point, multi-scale wavelet transform is performed to calculate the time-frequency coupling index of phase consistency and deformation energy. Feature frequency bands related to soil instability are screened, a phase coupling matrix is ​​constructed based on the feature frequency bands, and the monitoring area is divided into sub-regions using a community detection algorithm. The specific implementation of calculating the phase coupling index of each sub-region includes: The displacement time series data of the i-th monitoring point are obtained using the Morlet wavelet basis. Perform multi-scale continuous wavelet transform to obtain the time-frequency domain representation. Where s is the scale parameter, which is automatically optimized through maximum overlap discrete wavelet transform to satisfy... Less than the preset convergence threshold; Compute the phase consistency index PC(s,t) and deformation energy index E(s,t) at scale s: , Where J is the imaginary unit, Let be the phase angle of the i-th monitoring point at scale s and time t, and N be the total number of monitoring points in the monitoring area. Let i be the topological role weight of the i-th monitoring point. For modulo operation; Constructing time-frequency coupling index And select characteristic frequency bands to calculate the cumulative coupling strength at each scale s. ,in, At the start time, when When the value exceeds the set threshold and the duration exceeds the preset time, the frequency band corresponding to the judgment scale s is the characteristic frequency band related to the instability of the soil and rock mass. Based on the identified characteristic frequency bands, the phase time series of each monitoring point in that frequency band is extracted. Construct a weighted phase coupling matrix C, where the elements of the matrix are... Elements in the matrix The implementation involves first calculating the phase lock value between monitoring points i and j within the time window, i.e., calculating the complex exponential function for all time points t. The average value is used to introduce a spatial attenuation factor, and the spatial Euclidean distance between monitoring points i and j is calculated. And apply the exponential decay function Weighting is performed, where The spatial attenuation coefficient characterizes the spatial influence range of soil and rock deformation. It is ultimately obtained by multiplying the phase-locked value by the spatial attenuation factor. , characterizing the phase synchronization strength between monitoring points i and j considering the influence of spatial distance; The phase coupling matrix C is divided into sub-regions using a community detection algorithm to obtain... For each spatially continuous sub-region, calculate the phase coupling index of the k-th sub-region: ,in This represents the k-th sub-region. This is a modulo operation.

5. The data processing method for monitoring deformation of soil and rock mass according to claim 4, characterized in that, A deformation state entropy model is constructed by integrating the topological role weights of each monitoring point and the phase coupling index of each sub-region. The temporal evolution trajectory of entropy values ​​is analyzed. When the rate of change of entropy values ​​continuously exceeds the preset classification threshold, the corresponding level of deformation risk early warning mechanism is triggered. The implementation includes: for For the k-th region among the sub-regions, normalize the topological role weights of each monitoring point within it to obtain the probability distribution. Calculate the local deformation entropy of subregion k. ,in Let be the phase coupling index of subregion k; Based on the local deformation entropy of subregion k Obtain the global deformation state entropy value Establish an entropy time series and calculate the rate of change of entropy between adjacent time points. If three consecutive time points Warnings are issued when the value is less than 0 and below the set threshold.

Images