A tunnel seismic wave data and geological information fusion risk grading method

Abstract

The present application relates to the technical field of geological disaster early warning, in particular to a risk grading method of tunnel seismic wave data and geological information fusion, comprising the following steps: S1, collecting tunnel multi-temporal seismic wave amplitude, frequency response and propagation delay information, identifying abnormal response events and extracting their spatial propagation paths in combination with tunnel structure to generate seismic wave induced response event chain set; S2, fusing geological information such as drilling profile, geological structure line and lithology distribution to output structure chain type conduction map; S3, counting the activity, derived direction and time sequence trend of structure unit in the conduction map, calculating the trend activation factor to generate a risk level atlas with predictability and evolution ability. The present application realizes dynamic, accurate and predictable grading evaluation of tunnel structure risk by constructing seismic response event chain and structure conduction network, fusing multi-source geological information and introducing trend activation factor.

Description

Technical Field

[0001] This invention relates to the field of geological disaster early warning technology, and in particular to a risk classification method that integrates tunnel seismic wave data and geological information. Background Technology

[0002] When tunnel projects traverse complex geological areas, they face various geological hazard risks such as fault fracture zones, weak surrounding rock, and karst cavities. To ensure the safety of tunnel construction and operation, seismic wave monitoring equipment is often deployed to sense and analyze the response characteristics of the surrounding medium, identify potential abnormal response areas, and assist in geological risk prediction and structural safety assessment. With the improvement of monitoring accuracy and data density, how to extract key risk characteristics from a large amount of spatiotemporal dynamic response and effectively correlate them with geological structures has become the core challenge for identifying high-risk areas.

[0003] Existing seismic wave data analysis methods are mostly based on single-point anomalous waveform analysis or local wave velocity change inversion, lacking systematic modeling of the propagation relationship between seismic wave response events, making it difficult to characterize the causal chain of anomalous responses. In addition, the identification of structural risks still relies on static geological parameters or empirical formulas, which cannot dynamically reflect the cumulative impact of seismic response trends on structural units, and also cannot take into account both spatial hierarchical characteristics and temporal evolution laws, resulting in risk classification results that are lagging, limited, and crude. Summary of the Invention

[0004] This invention provides a risk classification method that integrates tunnel seismic wave data and geological information. It constructs a seismic wave-induced response event chain, integrates multi-source geological information to establish a structural response transmission mapping network, and drives classification rules by constructing trend activation factors. Finally, it outputs a risk level map with predictive and evolutionary capabilities, which can realize a closed loop of the entire process from seismic response anomaly identification to structural risk spatial mapping and dynamic assessment, thereby improving the accuracy and foresight of tunnel risk perception.

[0005] To achieve the above technical objectives, this invention provides a risk classification method for fusing tunnel seismic wave data and geological information, specifically including the following steps:

[0006] S1 collects multi-temporal seismic wave amplitude, frequency response, and propagation delay information in the tunnel monitoring area, identifies abnormal response events at different time points, and extracts the spatial propagation path of abnormal events in conjunction with the tunnel structure layout to generate a seismic wave induced response event chain set.

[0007] S2, based on the seismic wave induced response event chain set, integrate geological information from drilling profiles, geological structural lines and lithological distribution, construct a structural response transmission mapping network that reflects the interaction between abnormal events and structural units, and output a structural chain transmission diagram with spatial hierarchy and response direction;

[0008] S3. Based on the activity, derivation direction and temporal trend of each structural unit in the response path in the structural chain transmission diagram, calculate the trend activation factor of each structural unit and embed it as the dominant weight into the hierarchical rules to output a risk level map with predictive and dynamic evolution capabilities.

[0009] In a further technical solution of the present invention, step S1 specifically includes:

[0010] S11, a three-component seismograph array is deployed at key locations of the tunnel structure (arch, sidewall, and working face) to collect seismic wave data in real time for multiple time periods, including seismic wave amplitude, frequency response, and propagation delay information. The collected seismic wave data is preprocessed, including detrending, filtering, and normalization operations, and a standardized amplitude time series is output.

[0011] S12, based on the standardized amplitude time series, calculates the instantaneous energy mutation factor at each monitoring point as a local anomaly criterion, determines the abnormal event response point, and constructs the abnormal event feature vector of the abnormal event response point;

[0012] S13. Based on the tunnel structure layout and the relative positions of the three-component seismometers, a topological adjacency matrix is ​​constructed. All node pairs that satisfy the propagation relationship are arranged in chronological order to construct multiple spatial propagation paths. Finally, the set of all event paths that satisfy the spatial-temporal progression characteristics is defined as the seismic wave induced response event chain set.

[0013] In a preferred embodiment of the present invention, step S11 includes:

[0014] S111, three-component seismographs are deployed at the tunnel structure's arch, sidewalls, and face. A group of instruments is placed every 20–30m at the arch to capture reflected waves. Instruments are staggered every 10–20m along the sidewalls to collect shear waves (S-waves). Instruments are pre-positioned ahead of the tunnel face in the advancing section to obtain forward penetration wave information. Each deployment point is denoted as S111. ,in, The three-component seismograph is numbered. For the set of deployment points, the vertical component is acquired separately using each three-component seismograph. Horizontal component Horizontal and vertical components All collected data are processed according to the set time window. Store in batches to form the original observation dataset. ;

[0015] S112, for each component signal Perform linear detrending to eliminate baseline drift;

[0016] S113 applies a bandpass filter to the detrended signal to retain the effective vibration frequency band;

[0017] S114, perform zero-mean unit variance normalization on each signal component of each deployment point;

[0018] S115, based on the normalized signal, outputs a standardized amplitude time sequence. .

[0019] In a preferred embodiment of the present invention, step S12 includes:

[0020] S121, at each three-component seismograph deployment point At this point, based on the normalized amplitude time series of the output Calculate the instantaneous value of total energy at each time point. ;

[0021] S122, the instantaneous value of total energy at each monitoring point. By performing a first-order difference approximation, the instantaneous energy mutation factor is obtained. ;

[0022] S123, Set mutation threshold ,when When an abnormal response event is detected, it is considered that an abnormal response event has occurred at that moment. For each abnormal point, key parameters are extracted, including the event timestamp. Total energy peak Dominant frequency , magnitude of mutation;

[0023] S124, Construct an abnormal event feature vector based on the extracted key parameters. Finally, the feature vector set of all abnormal event points is output. .

[0024] In a preferred embodiment of the present invention, step S13 includes:

[0025] S131, based on the tunnel structure plan and the coordinates of the seismograph layout. Calculate the spatial Euclidean distance between any two three-component seismographs. Set a spatial adjacency threshold and define a spatial topological adjacency matrix. ;

[0026] S132, based on the feature vector of each abnormal event Extract its occurrence time and three-component seismograph location index For the feature vector of abnormal events , If there is a propagation path in space, the propagation is sequential in time, and the time difference does not exceed the set propagation window... Then it is believed yes The propagable successor point is denoted as And construct a time-ordered propagation graph structure. Its nodes are the feature vectors of all abnormal events, and its edges are the propagation paths;

[0027] S133, Based on Time-Ordered Propagation Graph Structure Starting from each anomalous event point, all seismic wave propagation paths that satisfy the space-time constraints are extracted, and all paths that satisfy the propagation conditions are used to form a seismic wave-induced response event chain set. .

[0028] In a preferred embodiment of the present invention, step S2 includes:

[0029] S21 divides geological tectonic lines and lithological distributions into several spatial structural units. Each structural unit To characterize a local region with independent geological properties, a spatial mapping matrix is ​​constructed for each seismic wave propagation path based on the matching results of the corresponding three-component seismograph positions and geological model coordinates. ;

[0030] S22, for each structural unit Extracting multi-source attribute vectors by fusing geological information And for each pair of structural units Transmission weights are constructed based on geological similarity and the direction of response influence. ;

[0031] S23, Construct a structural chain transmission diagram In this context, a node is a set of structural units. An edge represents a pair of structural units with non-zero propagation weights. The edge direction is driven by the event chain propagation path direction, and the edge weight is... .

[0032] In a preferred embodiment of the present invention, step S3 includes:

[0033] S31, based on the structural chain transmission diagram, statistically analyze the frequency of each structural unit in the response path, the proportion of propagation direction, and the trend of response time change, and comprehensively form a trend activation factor;

[0034] S32, based on the magnitude of the trend activation factor, divides the structural units into different risk levels, and combines the structural topological relationship to expand and adjust the adjacent areas, generating a risk level map with spatial hierarchy and evolution capabilities.

[0035] In a preferred embodiment of the present invention, step S31 includes:

[0036] S311, for each structural unit in the structural chain transmission diagram Traverse all seismic wave-induced response event chains and statistically analyze structural units. Number of times And calculate its normalized active frequency factor. ;

[0037] S312, For each seismic wave propagation path, based on its structural unit mapping sequence Statistical structural unit The propagation directionality factor of the structural unit is calculated by considering the number of times the starting and ending nodes are used. ;

[0038] S313, Extract all seismic wave-induced response event chains and map them to structural elements. event set Calculate the mean square slope of its response time series. ;

[0039] S314, for the normalized active frequency factor , propagation direction factor and mean square slope Weighted fusion is performed to generate trend activation factors. .

[0040] In a preferred embodiment of the present invention, step S32 includes:

[0041] S321, Activation factor based on the trend of all structural units Calculate the mean of all values. with standard deviation Construct a multi-level risk classification range, including low risk, medium risk, high risk, and extremely high risk, and for each structural unit Assign risk labels Among them, 0 represents low risk, 1 represents medium risk, 2 represents high risk, and 3 represents extremely high risk;

[0042] S322, using structural topological adjacency matrix Represent the spatial adjacency relationship between structural units, define a risk propagation smoothing function, and calculate the risk level of the structural units after smoothing. ;

[0043] S323, based on the spatial position of structural units Coordinates, risk level labels Using the map values, generate a three-dimensional distributed risk level map. .

[0044] The preferred technical solution of this invention: the construction of multi-level risk classification intervals in step S321 includes:

[0045] when At that time, it was classified as low risk;

[0046] when At that time, it was classified as medium risk;

[0047] when At that time, it was classified as high-risk;

[0048] when At that time, it was classified as extremely high risk.

[0049] The beneficial effects of this invention are:

[0050] This invention constructs a multi-temporal anomaly identification mechanism based on seismic wave response events, combines a three-component seismograph deployment method with an energy mutation factor calculation method, achieves high-precision extraction of anomaly response events in complex geological environments, and utilizes spatial topological constraints to construct a seismic wave-induced response event chain set, significantly improving the spatiotemporal continuity and geological relevance of anomaly event identification.

[0051] This invention integrates drilling profiles, geological structural lines, and lithological distribution information to construct a structural response transmission diagram that reflects the interaction between abnormal events and structural units. Furthermore, it constructs a structural chain response network through the similarity of geological attributes and the directionality of propagation, thereby realizing the dynamic mapping from the abnormal event chain to the geological structural level. This effectively solves the problem of the disconnect between spatial structure and dynamic response in existing technologies.

[0052] This invention constructs a trend activation factor and introduces the activity, directionality, and response time-series trend characteristics of structural units to realize a dynamic risk classification mechanism oriented towards the evolution of seismic response. Combined with a spatial topology expansion strategy, it generates a risk level map with hierarchy and evolution, which not only improves the predictive ability of risk assessment, but also provides highly adaptable technical support for disaster prevention and control during tunnel construction and operation. Attached Figure Description

[0053] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only for this invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0054] Figure 1 This is a schematic diagram of the grading method according to an embodiment of the present invention;

[0055] Figure 2 This is a schematic diagram illustrating the construction of a seismic wave-induced response event chain set according to an embodiment of the present invention. Detailed Implementation

[0056] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. Those skilled in the art may employ other alternative methods to implement some well-known technologies; moreover, the accompanying drawings are only for more specific description of the embodiments and are not intended to specifically limit the present invention.

[0057] The embodiment provides a risk classification method that fuses tunnel seismic wave data with geological information, such as Figures 1-2 As shown, it includes the following steps:

[0058] S1. Collect multi-temporal seismic wave amplitude, frequency response, and propagation delay information in the tunnel monitoring area; identify anomalous response events at different time points; and, in conjunction with the tunnel structure layout, extract the spatial propagation paths of the anomalous events to generate a seismic wave-induced response event chain set; specifically including the following steps:

[0059] S11, a three-component seismograph array is deployed at key locations in the tunnel structure (arch, sidewalls, and tunnel face) to acquire seismic wave data in real time across multiple time periods, including seismic wave amplitude, frequency response, and propagation delay information. The acquired seismic wave data is preprocessed, including detrending, filtering, and normalization operations, and a standardized amplitude time series is output; details are as follows:

[0060] S111, three-component seismographs are deployed at the tunnel's arch, sidewalls, and face. One set of instruments is placed every 20-30m at the arch to capture reflected waves. Instruments are staggered every 10-20m along the sidewalls to collect shear waves (S-waves). Instruments are pre-positioned ahead of the tunnel face in the advancing section to obtain penetrating wave information. Each deployment point is denoted as [instrument name missing]. ,in, The three-component seismograph is numbered. For the set of deployment points, the vertical component is acquired separately using each three-component seismograph. Horizontal component Horizontal and vertical components All collected data are processed according to the set time window. Store in batches to form the original observation dataset. ;

[0061] S112, for each component signal Perform linear detrending to eliminate baseline drift, as shown below:

[0062] ;

[0063] in, , These represent the slope and intercept of the linear fit for the corresponding components. This is a signal after detrending;

[0064] S113 applies a bandpass filter to the detrended signal to retain the effective vibration frequency band, as shown below:

[0065] ;

[0066] in, The filtered signal This indicates a standard digital bandpass filter operation. , ;

[0067] S114, performs zero-mean unit variance normalization on each signal component at each deployment point, expressed as:

[0068] ;

[0069] in, , These are the mean and standard deviation of the filtered signal, respectively. The signal is standardized.

[0070] S115, based on the normalized signal, outputs a standardized amplitude time sequence. , represented as:

[0071] .

[0072] S12, based on standardized amplitude time series, calculates the instantaneous energy mutation factor at each monitoring point as a local anomaly criterion, determines the abnormal event response point, and constructs the abnormal event feature vector of the abnormal event response point; specifically including:

[0073] S121, at each three-component seismograph deployment point At this point, based on the normalized amplitude time series of the output Calculate the instantaneous value of total energy at each time point. , represented as:

[0074] ;

[0075] in, , , The first , , The normalized amplitude of the component;

[0076] S122, the instantaneous value of total energy at each monitoring point. By performing a first-order difference approximation, the instantaneous energy mutation factor is obtained. , represented as:

[0077] ;

[0078] in, The sampling time interval;

[0079] S123, Set mutation threshold ,when When an abnormal response event is detected, it is considered that an abnormal response event has occurred at that moment. For each abnormal point, key parameters are extracted, including the event timestamp. Total energy peak Dominant frequency Mutation amplitude ;

[0080] Mutation threshold Represented as:

[0081] ;

[0082] in, This represents the abnormal activation sensitivity coefficient. , These are the mean and standard deviation of the mutation factor, respectively.

[0083] S124, Construct an abnormal event feature vector based on the extracted key parameters. Finally, the feature vector set of all abnormal event points is output. , represented as:

[0084] ;

[0085] .

[0086] S13. Based on the tunnel structure layout and the relative positions of the three-component seismometers, a topological adjacency matrix is ​​constructed. For all node pairs satisfying the propagation relationship, they are arranged in chronological order to construct multiple spatial propagation paths. Finally, the set of all event paths satisfying the spatial-temporal progression characteristics is defined as the seismic wave-induced response event chain set; specifically including:

[0087] S131, based on the tunnel structure plan and the coordinates of the seismograph layout. Calculate the spatial Euclidean distance between any two three-component seismographs. Set a spatial adjacency threshold (30m) and define a spatial topological adjacency matrix. , represented as:

[0088] ;

[0089] ;

[0090] S132, based on the feature vector of each abnormal event Extract its occurrence time and three-component seismograph location index For the feature vector of abnormal events , (Corresponding deployment points) , If there is a propagation path in space (i.e.) ), and satisfy the sequentiality in time (i.e. And the time difference does not exceed the set propagation window. ( ), then it is believed yes The propagable successor point is denoted as And construct a time-ordered propagation graph structure. Its nodes are the feature vectors of all abnormal events, and its edges are the propagation paths;

[0091] S133, Based on Time-Ordered Propagation Graph Structure Starting from each anomalous event point, extract all seismic wave propagation paths that satisfy the space-time constraints. Each path takes the following form:

[0092] ;

[0093] in, For the first A chain of seismic wave-induced response events, For the first in the event chain The feature vector of a response event , This represents the total number of response events contained in the event chain. For the first The timestamp of each response event;

[0094] All paths that satisfy the propagation conditions are used to form a seismic wave induced response event chain set. , represented as:

[0095] ;

[0096] in, The total number of paths required to satisfy the propagation conditions.

[0097] S2, based on the seismic wave-induced response event chain set, integrates geological information from drilling profiles, geological structural lines, and lithological distribution to construct a structural response transmission mapping network reflecting the interaction between anomalous events and structural units, outputting a structural chain transmission diagram with spatial hierarchy and response direction; the S2 step specifically includes the following steps:

[0098] S21 divides geological tectonic lines and lithological distributions into several spatial structural units. Each structural unit To characterize a local region with independent geological properties (such as fault segments, karst zones, weak zones, etc.), a spatial mapping matrix is ​​constructed for each seismic wave propagation path based on the matching results of the corresponding three-component seismograph positions and geological model coordinates. , represented as:

[0099] ;

[0100] S22, for each structural unit Extracting multi-source attribute vectors by fusing geological information And for each pair of structural units Transmission weights are constructed based on geological similarity and the direction of response influence. , represented as:

[0101] ;

[0102] ;

[0103] in, It is a lithological density index (rock type code, density grade). Weights for fault influence (such as the number of fault intersections and the distance from the main fault). This refers to the proportion of drilling anomalies (such as the proportion of abnormal rock layers in the drilling profile). As a directional indicator factor (such as determining the direction of propagation based on chain concentration) (Whether it is reasonable; if reasonable, the value is 1; otherwise, the value is 0).

[0104] S23, Construct a structural chain transmission diagram In this context, a node is a set of structural units. An edge represents a pair of structural units with non-zero propagation weights. The edge direction is driven by the event chain propagation path direction, and the edge weight is... .

[0105] S3, based on the activity, derivation direction, and temporal trend of each structural unit in the response path of the structural chain transmission diagram, calculates the trend activation factor of each structural unit and embeds it as the dominant weight into the hierarchical rules, outputting a risk level map with predictive and dynamic evolution capabilities; Step S3 specifically includes:

[0106] S31, based on the structural chain transmission diagram, statistically analyze the frequency of each structural unit in the response path, the proportion of its propagation direction, and the trend of its response time. This data is then used to form a trend activation factor, which measures the activity and dynamic evolution tendency of structural units in seismic wave response, reflecting their criticality and sensitivity in the abnormal event chain. Detailed steps are as follows:

[0107] S311, for each structural unit in the structural chain transmission diagram Traverse all seismic wave-induced response event chains and statistically analyze structural units. Number of times And calculate its normalized active frequency factor. , represented as:

[0108] ;

[0109] S312, For each seismic wave propagation path, based on its structural unit mapping sequence Statistical structural unit The propagation directionality factor of the structural unit is calculated by considering the number of times the starting and ending nodes are used. , represented as:

[0110] ;

[0111] in, For the path with The number of times starting from the point of view For the path with The number of times the endpoint is reached. If it is a local minimum, If the value approaches 1, it indicates that the structural unit is mostly a propagation source; if it approaches 0, it indicates that it is mostly a path endpoint.

[0112] S313, Extract all seismic wave-induced response event chains and map them to structural elements. event set Calculate the mean square slope of its response time series. (Trend strength), expressed as:

[0113] ;

[0114] ;

[0115] in, , These are the numbers sorted by time. , Response time The distance of the propagation path between structural units;

[0116] S314, for the normalized active frequency factor , propagation direction factor and mean square slope Weighted fusion is performed to generate trend activation factors. , represented as:

[0117] ;

[0118] in, , , These are the corresponding weighting coefficients.

[0119] S32, based on the numerical value of the trend activation factor, divides structural units into different risk levels, and expands and adjusts adjacent areas in conjunction with structural topological relationships to generate a risk level map with spatial hierarchy and evolutionary capabilities. This map is used to predict potentially high-risk areas and improve the risk early warning capability for tunnel construction or operation; specifically including:

[0120] S321, Activation factor based on the trend of all structural units Calculate the mean of all values. with standard deviation Construct a multi-level risk classification range, including low risk, medium risk, high risk, and extremely high risk, and for each structural unit Assign risk labels Where 0 represents low risk, 1 represents medium risk, 2 represents high risk, and 3 represents extremely high risk, as shown below:

[0121] ;

[0122] ;

[0123] in, This represents the total number of trend activating factors.

[0124] S322, using structural topological adjacency matrix Represent the spatial adjacency relationship between structural units, define a risk propagation smoothing function, and calculate the risk level of the structural units after smoothing. , represented as:

[0125] ;

[0126] in, To and Adjacent structural unit sets, This is the risk transmission coefficient;

[0127] S323, based on the spatial position of structural units Coordinates, risk level labels Using the map values, generate a three-dimensional distributed risk level map. .

[0128] Constructing a multi-level risk grading range includes:

[0129] when At that time, it was classified as low risk;

[0130] when At that time, it was classified as medium risk;

[0131] when At that time, it was classified as high-risk;

[0132] when At that time, it was classified as extremely high risk.

[0133] This invention encompasses any substitutions, modifications, equivalent methods, and solutions made within the spirit and scope of this invention. To provide the public with a thorough understanding of this invention, specific details are described in detail in the following preferred embodiments; however, those skilled in the art will fully understand the invention even without these details. Furthermore, to avoid unnecessary misunderstanding of the essence of this invention, well-known methods, processes, procedures, components, and circuits are not described in detail.

[0134] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A risk classification method that integrates tunnel seismic wave data and geological information, characterized in that, Specifically, the following steps are included: S1 collects multi-temporal seismic wave amplitude, frequency response, and propagation delay information in the tunnel monitoring area, identifies abnormal response events at different time points, and extracts the spatial propagation path of abnormal events in conjunction with the tunnel structure layout to generate a seismic wave induced response event chain set. S2, based on the seismic wave induced response event chain set, integrate geological information from drilling profiles, geological structural lines and lithological distribution, construct a structural response transmission mapping network that reflects the interaction between abnormal events and structural units, and output a structural chain transmission diagram with spatial hierarchy and response direction; S3. Based on the activity, derivation direction and temporal trend of each structural unit in the response path in the structural chain transmission diagram, calculate the trend activation factor of each structural unit and embed it as the dominant weight into the hierarchical rules to output a risk level map with predictive and dynamic evolution capabilities.

2. The risk classification method for fusing tunnel seismic wave data and geological information according to claim 1, characterized in that, Step S1 includes: S11, a three-component seismograph array is deployed at key locations in the tunnel structure to collect seismic wave data in real time for multiple time periods, including seismic wave amplitude, frequency response and propagation delay information, and preprocess the collected seismic wave data, including detrending, filtering and normalization operations, and output standardized amplitude time series. S12, based on the standardized amplitude time series, calculates the instantaneous energy mutation factor at each monitoring point as a local anomaly criterion, determines the abnormal event response point, and constructs the abnormal event feature vector of the abnormal event response point; S13. Based on the tunnel structure layout and the relative positions of the three-component seismometers, a topological adjacency matrix is ​​constructed. All node pairs that satisfy the propagation relationship are arranged in chronological order to construct multiple spatial propagation paths. Finally, the set of all event paths that satisfy the spatial-temporal progression characteristics is defined as the seismic wave induced response event chain set.

3. The risk classification method for fusing tunnel seismic wave data and geological information according to claim 2, characterized in that, Step S11 includes: S111, three-component seismographs are deployed at the tunnel structure's arch, sidewalls, and face. A group of instruments is placed every 20-30m at the arch to capture reflected waves. Instruments are staggered every 10-20m along the sidewalls to collect shear waves. Instruments are pre-deployed ahead of the tunnel face in the advancing section to obtain forward penetration wave information. Each deployment point is denoted as S111. ,in, The three-component seismograph is numbered. For the set of deployment points, the vertical component is acquired separately using each three-component seismograph. Horizontal component Horizontal and vertical components All collected data are processed according to the set time window. Store in batches to form the original observation dataset. ; S112, for each component signal Perform linear detrending to eliminate baseline drift; S113 applies a bandpass filter to the detrended signal to retain the effective vibration frequency band; S114, perform zero-mean unit variance normalization on each signal component of each deployment point; S115, based on the normalized signal, outputs a standardized amplitude time sequence. .

4. The risk classification method for fusing tunnel seismic wave data and geological information according to claim 3, characterized in that, Step S12 includes: S121, at each three-component seismograph deployment point At this point, based on the normalized amplitude time series of the output Calculate the instantaneous value of total energy at each time point. ; S122, the instantaneous value of total energy at each monitoring point. By performing a first-order difference approximation, the instantaneous energy mutation factor is obtained. ; S123, Set mutation threshold ,when When an abnormal response event is detected, it is considered that an abnormal response event has occurred at that moment. For each abnormal point, key parameters are extracted, including the event timestamp. Total energy peak Dominant frequency , magnitude of mutation; S124, Construct an abnormal event feature vector based on the extracted key parameters. Finally, the feature vector set of all abnormal event points is output. .

5. The risk classification method for fusing tunnel seismic wave data and geological information according to claim 4, characterized in that, Step S13 includes: S131, based on the tunnel structure plan and the coordinates of the seismograph layout. Calculate the spatial Euclidean distance between any two three-component seismographs. Set a spatial adjacency threshold and define a spatial topological adjacency matrix. ; S132, based on the feature vector of each abnormal event Extract its occurrence time and three-component seismograph location index For the feature vector of abnormal events , If there is a propagation path in space, the temporal order is satisfied, and the time difference does not exceed the set propagation window. Then it is believed yes The propagable successor point is denoted as And construct a time-ordered propagation graph structure. Its nodes are the feature vectors of all abnormal events, and its edges are the propagation paths; S133, Based on Time-Ordered Propagation Graph Structure Starting from each anomalous event point, all seismic wave propagation paths that satisfy the space-time constraints are extracted, and all paths that satisfy the propagation conditions are used to form a seismic wave-induced response event chain set. .

6. The risk classification method for fusing tunnel seismic wave data and geological information according to claim 5, characterized in that, Step S2 includes: S21 divides geological tectonic lines and lithological distributions into several spatial structural units. Each structural unit To characterize a local region with independent geological properties, a spatial mapping matrix is ​​constructed for each seismic wave propagation path based on the matching results of the corresponding three-component seismograph positions and geological model coordinates. ; S22, for each structural unit Extracting multi-source attribute vectors by fusing geological information And for each pair of structural units Transmission weights are constructed based on geological similarity and the direction of response influence. ; S23, Construct a structural chain transmission diagram In this context, a node is a set of structural units. An edge represents a pair of structural units with non-zero propagation weights. The edge direction is driven by the event chain propagation path direction, and the edge weight is... .

7. The risk classification method for fusing tunnel seismic wave data and geological information according to claim 6, characterized in that, Step S3 includes: S31, based on the structural chain transmission diagram, statistically analyze the frequency of each structural unit in the response path, the proportion of propagation direction, and the trend of response time change, and comprehensively form a trend activation factor; S32, based on the magnitude of the trend activation factor, divides the structural units into different risk levels, and combines the structural topological relationship to expand and adjust the adjacent areas, generating a risk level map with spatial hierarchy and evolution capabilities.

8. The risk classification method for fusing tunnel seismic wave data and geological information according to claim 7, characterized in that, Step S31 includes: S311, for each structural unit in the structural chain transmission diagram Traverse all seismic wave-induced response event chains and statistically analyze structural units. Number of times And calculate its normalized active frequency factor. ; S312, For each seismic wave propagation path, based on its structural unit mapping sequence Statistical structural unit The propagation directionality factor of the structural unit is calculated by considering the number of times the starting and ending nodes are used. ; S313, Extract all seismic wave-induced response event chains and map them to structural elements. event set Calculate the mean square slope of its response time series. ; S314, for the normalized active frequency factor , propagation direction factor and mean square slope Weighted fusion is performed to generate trend activation factors. .

9. The risk classification method for fusing tunnel seismic wave data and geological information according to claim 8, characterized in that, Step S32 includes: S321, Activation factor based on the trend of all structural units Calculate the mean of all values. with standard deviation Construct a multi-level risk classification range, including low risk, medium risk, high risk, and extremely high risk, and for each structural unit Assign risk labels Among them, 0 represents low risk, 1 represents medium risk, 2 represents high risk, and 3 represents extremely high risk; S322, using structural topological adjacency matrix Represent the spatial adjacency relationship between structural units, define a risk propagation smoothing function, and calculate the risk level of the structural units after smoothing. ; S323, based on the spatial position of structural units Coordinates, risk level labels Using the map values, generate a three-dimensional distributed risk level map. .

10. The risk classification method for fusing tunnel seismic wave data and geological information according to claim 9, characterized in that, The construction of multi-level risk grading intervals in step S321 includes: when At that time, it was classified as low risk; when At that time, it was classified as medium risk; when At that time, it was classified as high-risk; when At that time, it was classified as extremely high risk.

Images