A method for detecting and predicting risks of poor geological bodies in a power transmission and distribution line corridor

Abstract

This invention relates to a method for detecting and predicting risks of adverse geological features in power transmission and distribution line corridors. The specific steps include: collecting frequency distribution characteristics of seismic wave signal data from the power transmission and distribution line corridor area; obtaining shallow geological body boundary information based on the frequency distribution characteristics of the seismic wave signals; performing multi-scale decomposition of the shallow geological body boundary information using wavelet transform to obtain a multi-scale wave coefficient sequence; determining the location of shallow faults and folds using a support vector machine classifier; obtaining a deep geological body distribution map based on the shallow fault and fold location data and low-frequency signal component characteristics; obtaining the location coordinates of potential hazards by combining the shallow fault and fold locations; simultaneously determining the propagation interference zone using the hazard location coordinates; generating an enhanced resolution signal by adjusting the weights of the frequency bands in the propagation interference zone of the seismic wave signals; reconstructing the overall geological model based on the enhanced resolution signal; determining the interaction between geological hazards and the power transmission line corridor in the model; and obtaining the final risk prediction result.

Description

Technical Field

[0001] This invention relates to the field of adverse geological body detection technology for power transmission and distribution line corridors, specifically a method for detecting and predicting risks of adverse geological bodies in power transmission and distribution line corridors. Background Technology

[0002] In modern infrastructure construction, the safety of power transmission and distribution line corridors is of paramount importance, directly impacting the stability of power supply and public safety. Especially in complex geological environments, the presence of adverse geological features such as faults and folds can pose serious threats to the planning and operation of power lines. Therefore, accurately identifying and locating these geological hazards has become an urgent need to ensure the safe operation of power facilities. Research in this field is not only crucial for engineering and technological advancements but also closely related to socio-economic stability.

[0003] However, current methods for detecting geological hazards along power transmission and distribution line corridors still have significant shortcomings. Many traditional methods are often ill-suited to complex geological environments and varying geological structures, particularly in distinguishing the boundaries of different types of geological bodies and detecting deep structures, easily leading to misjudgments or omissions. This limitation mainly stems from insufficient comprehensive analytical capabilities regarding geological body characteristics, resulting in an inability to accurately identify potential risk points in practical engineering projects.

[0004] Looking further, the core technical challenge in this field lies in effectively capturing and analyzing the propagation characteristics of seismic waves in complex geological environments. As a crucial tool for detecting underground structures, seismic waves are affected by various factors during their propagation in different geological bodies, particularly the differences in frequency components. These differences directly impact the wave's penetration ability and resolution. High-frequency waves may provide better detail information in shallow layers but struggle to penetrate deeper structures, while low-frequency waves, although capable of detecting deeper areas, are prone to losing subtle features. This contradiction in frequency characteristics makes it difficult to simultaneously and accurately identify both deep structures and shallow details in practical exploration.

[0005] Therefore, effectively analyzing the differences in seismic wave frequency components within complex geological formations during the detection of power transmission and distribution line corridors, and balancing the needs of deep detection and shallow resolution through appropriate technical means, has become a key issue in ensuring the planning and operational safety of power transmission lines. Solving this problem will directly impact the scientific validity of project site selection and the reliability of long-term operation. Summary of the Invention

[0006] To address the problems existing in the prior art, this invention proposes a method for detecting and predicting risks of adverse geological features in power transmission and distribution line corridors.

[0007] The technical solution of the present invention is as follows: On the one hand, this invention proposes a method for detecting and predicting risks of adverse geological features in power transmission and distribution line corridors, the specific steps of which include: Seismic wave signal data were collected from the power transmission and distribution line corridor area. High-frequency and low-frequency signal components were separated using time-frequency analysis to obtain frequency distribution characteristics. Based on the frequency distribution characteristics of the seismic wave signals, the boundary information of shallow geological bodies was obtained. Wavelet transform was used to decompose the boundary information of shallow geological bodies at multiple scales to obtain a multi-scale wave coefficient sequence. The peak features of the multi-scale wave coefficient sequence were extracted using a threshold filtering method. The peak features were then matched with a pre-established fault pattern library using a support vector machine classifier to determine the location of shallow fault folds. Based on the location data of shallow faults and folds and the characteristic information of low-frequency signal components, an inversion algorithm is used to reconstruct the deep structural model and obtain the distribution map of deep geological bodies. High-risk geological hazard areas are identified by the overlap between the distribution map of deep geological bodies and the location of shallow faults and folds, and the coordinates of the hazards are obtained. At the same time, the areas of propagation interference are determined by the coordinates of the hazards. An enhanced resolution signal is generated by adjusting the weights of the frequency bands in the interference zone of the seismic wave signal. The overall geological model is reconstructed based on the enhanced resolution signal, and the interaction between geological hazards and the route corridor in the model is determined to obtain the final risk prediction results.

[0008] In a preferred embodiment, the step of obtaining shallow geological body boundary information based on the frequency distribution characteristics of seismic wave signals specifically includes: Calculate the energy ratio of high-frequency signal components to low-frequency signal components based on frequency distribution characteristics; If the energy ratio is greater than the preset threshold, it is determined to be a shallow detail-dominated region, and the boundary information of the shallow geological body in that region is extracted.

[0009] As a preferred embodiment, the step of using wavelet transform to perform multi-scale decomposition of shallow geological body boundary information and obtaining a multi-scale wave coefficient sequence specifically includes: Wavelet transform was used to decompose the boundary information of shallow geological bodies at multiple scales to obtain a preliminary wave coefficient sequence. Key boundary features are marked based on signal intensity at different scales; The preliminary wave coefficient sequence is reorganized based on key boundary features to obtain an optimized multi-scale wave coefficient sequence.

[0010] In a preferred embodiment, the step of matching peak features with a pre-established fault pattern library using a support vector machine classifier to determine the location of shallow fault folds specifically involves: The peak features are input into a support vector machine classifier and matched with a pre-set tomographic pattern library to obtain the classification result. Then, the location of shallow fault folds is determined based on the classification results.

[0011] As a preferred embodiment, the step of reconstructing a deep structural model based on shallow fault and fold location data and low-frequency signal component feature information using an inversion algorithm to obtain a deep geological body distribution map specifically includes: Extract relevant low-frequency signal components based on the location of shallow fault folds; Extracting key signal data related to deep structures from low-frequency signal components; Based on key signal data, a deep structural model is generated through an inversion algorithm, and a distribution map of deep geological bodies is obtained by combining it with geological mapping technology.

[0012] As a preferred embodiment, the step of determining high-risk geological hazard areas based on the overlap between the distribution map of deep geological bodies and the location of shallow faults and folds specifically includes: The distribution map of deep geological bodies is spatially overlaid with the location of shallow faults and folds, and the percentage of overlapping areas is calculated. If the percentage of overlapping areas is greater than a preset threshold, it is determined to be a high-risk geological hazard area.

[0013] In a preferred embodiment, the step of determining the propagation interference zone using the hazard location coordinates specifically includes: A three-dimensional mesh structure is constructed based on the coordinates of the hazard location, and then seismic wave propagation parameters are loaded based on the three-dimensional mesh structure; Numerical simulations were performed using the finite difference method to extract the propagation trajectory of seismic waves. Grid cells whose deviation from the seismic wave propagation trajectory exceeds a preset threshold are identified as propagation interference zones.

[0014] In a preferred embodiment, the step of generating an enhanced resolution signal by adjusting the weights of the frequency bands in the propagation interference zone of the seismic wave signal specifically includes: For the signal frequency range of the propagation interference zone, the frequency ranges of the propagation interference zone and the non-propagation interference zone are distinguished from the frequency range of the seismic wave signal to obtain the initial set of seismic wave signal frequency components. Set the weights for each frequency range of the initial seismic wave signal frequency component set, and resynthesize and optimize the seismic wave signal through weighted processing; An adaptive filter is used to perform secondary filtering on the optimized seismic wave signal to obtain the filtered signal components. The filtered signal components are then enhanced to obtain a signal with enhanced resolution.

[0015] On the other hand, the present invention proposes an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements a method for detecting and predicting adverse geological features in a power transmission and distribution line corridor as described in any embodiment of the present invention.

[0016] On the other hand, the present invention proposes a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements a method for detecting and predicting adverse geological features in a power transmission and distribution line corridor as described in any embodiment of the present invention.

[0017] The present invention has the following beneficial effects: 1. This invention separates the high and low frequency components of seismic wave signals through time-frequency analysis and determines shallow detailed areas by combining energy ratios. This effectively solves the technical contradiction of traditional methods that make it difficult to take into account both deep and shallow geological features, and significantly improves the accuracy of identifying the boundaries of complex geological bodies.

[0018] 2. This invention employs wavelet transform multi-scale decomposition and threshold filtering to extract peak features, and combines a support vector machine classifier with a fault pattern library for intelligent matching. This enables rapid and accurate identification of the location and morphological features of shallow fault folds, reducing the risk of misjudgment.

[0019] 3. This invention overcomes the technical difficulties of weak deep signals and numerous interferences by fusing shallow fault and fold data with low-frequency signal components and using an inversion algorithm to reconstruct a deep structural model and generate a high-resolution distribution map of deep geological bodies.

[0020] 4. This invention analyzes the overlap of deep and shallow geological data through spatial overlay, automatically identifies high-risk areas based on preset thresholds, and generates hazard location coordinates, thus achieving a leap from qualitative judgment to quantitative assessment and improving the reliability of risk decision-making.

[0021] 5. This invention constructs a three-dimensional grid to simulate the propagation path of seismic waves, identifies the propagation interference zone, and adjusts the frequency weights accordingly. Combined with adaptive filtering and waveform enhancement techniques, it effectively suppresses noise interference and generates a high signal-to-noise ratio enhanced resolution signal.

[0022] 6. This invention is based on the reconstruction of the overall geological model by enhanced signal, and comprehensively analyzes the interaction between geological hazards and line corridors. It realizes the integration of the entire process from data acquisition, processing, modeling to risk prediction, and provides complete technical support for the planning and safe operation and maintenance of power transmission and distribution lines. Attached Figure Description

[0023] Figure 1 This is a flowchart of the steps of the present invention. Detailed Implementation

[0024] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0025] It should be understood that the step numbers used in the text are for ease of description only and are not intended to limit the order in which the steps are performed.

[0026] It should be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0027] The terms “comprising” and “including” indicate the presence of the described feature, whole, step, operation, element and / or component, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or collections thereof.

[0028] The term “and / or” refers to any combination of one or more of the associated listed items, as well as all possible combinations, and includes these combinations.

[0029] Example 1: See Figure 1 A method for detecting and predicting risks of adverse geological features in power transmission and distribution line corridors, comprising the following steps: Step S101: Collect seismic wave signal data in the power transmission and distribution line corridor area, and use time-frequency analysis to separate high-frequency and low-frequency signal components to obtain frequency distribution characteristics; obtain shallow geological body boundary information based on the frequency distribution characteristics of the seismic wave signals. Data was collected from the power transmission and distribution line corridor area using seismic exploration equipment to obtain raw seismic wave signal data, which was stored as an initial signal dataset. Based on this dataset, time-frequency analysis was used to decompose the signal, separating high-frequency and low-frequency components to obtain a subset of the decomposed signal. For this subset, the frequency distribution characteristics of the high-frequency and low-frequency components were extracted to determine frequency distribution parameters. If the proportion of high-frequency components in the frequency distribution parameters exceeded a preset threshold, further signal filtering was applied to the high-frequency components to obtain a filtered high-frequency signal subset. The filtered high-frequency signal subset was compared with the low-frequency component signal subset to determine if any anomalous frequency distribution intervals existed. Based on the determination of anomalous frequency distribution intervals and combined with geological data from the corridor area, potential geological anomaly locations were identified. A pre-established geological anomaly location database was used to match and verify the potential geological anomaly locations, yielding the final anomaly location distribution results.

[0030] Specifically, seismic wave signal data of the power transmission and distribution line corridor area is collected using seismic exploration equipment. In practice, high-precision seismic wave sensors are deployed at multiple measuring points within the target area, for example, one measuring point every 100 meters, covering a total of 10 kilometers of the corridor. Seismic wave signals with a frequency range of 0.5 Hz to 200 Hz are collected at a sampling rate of 1000 Hz to ensure the capture of weak vibration signals. The data is stored as a time-series waveform file. Subsequently, time-frequency analysis methods are used to separate high-frequency and low-frequency components. Wavelet transform algorithms are used to decompose the signal into multiple frequency bands, setting the threshold for the low-frequency component to 0.5 to 20 Hz and the threshold for the high-frequency component to 20 to 200 Hz. Discrete wavelet transform (DWT) is used to perform multi-scale decomposition of the signal, extracting the coefficients of each frequency band and reconstructing the signal to obtain the separated low-frequency and high-frequency waveform data. The energy distribution is then analyzed; for example, the low-frequency component accounts for 60% of the energy, and the high-frequency component accounts for 40%. Finally, when obtaining frequency distribution characteristics, the separated signals are analyzed by Fast Fourier Transform (FFT) to calculate the power spectral density, identify the dominant frequency of the low-frequency band as 5 Hz and the dominant frequency of the high-frequency band as 50 Hz, and observe the concentration and bandwidth characteristics of the frequency distribution in conjunction with the power spectrum diagram. For example, the low-frequency bandwidth is 2 Hz and the high-frequency bandwidth is 10 Hz. At the same time, the operating status data of the transmission and distribution lines, such as the frequency of current load change, can be correlated. It was found that the low-frequency signal is highly correlated with the load fluctuation frequency of 5 Hz, which suggests that it may be affected by the mechanical vibration of the line, while the high-frequency signal may be related to external interference such as wind vibration.

[0031] Raw geological signal data is acquired using signal acquisition equipment. The time-domain signal is transformed into a frequency-domain signal using a frequency domain conversion method to obtain the frequency distribution results. Based on the frequency distribution results, high-frequency and low-frequency components are separated, and their energy values ​​are calculated using energy calculation formulas to determine the energy ratio. If the energy ratio is greater than a preset threshold, the corresponding area is identified as a shallow detail-dominated area, and a preliminary dominant area distribution is obtained. Spatial mapping processing is performed on the dominant area distribution to extract the contour data of the geological body edges, resulting in a preliminary boundary delineation. Local signal enhancement processing is performed using the boundary information data to highlight shallow detail features and determine a more precise boundary range. Based on the precise boundary range data and combined with regional determination logic, the geological body edges are classified and labeled to obtain the final shallow geological body boundary distribution results. Data storage and visualization processing of the final boundary distribution results generate a structured boundary information archive, completing the comprehensive analysis of the shallow geological body boundaries.

[0032] Specifically, in processing geological data to calculate the energy ratio of high-frequency to low-frequency components and determine the dominant shallow detail region, the first step is to perform a Fourier transform on the acquired geological signal data, decomposing it into different components in the frequency domain. Assuming the input signal is a data sequence containing 1024 sampling points at a sampling frequency of 1000 Hz, after a Fast Fourier Transform, a spectral distribution with a frequency range from 0 to 500 Hz is obtained. Next, a frequency threshold of above 200 Hz is set for the high-frequency components, and below 200 Hz for the low-frequency components. The energy values ​​within each frequency range are calculated separately. The specific algorithm involves integrating the sum of squared amplitudes at each frequency point in the spectrum. The high-frequency energy value is 1200 units, and the low-frequency energy value is 800 units. The energy ratio is 1200 divided by 800, which equals 1.5. This ratio is then compared to a preset threshold of 1.2. Finding that 1.5 is greater than 1.2, the region is determined to be a dominant shallow detail region. To further obtain boundary information of shallow geological bodies, an edge detection algorithm was employed, such as multi-scale analysis based on wavelet transform. The wavelet basis was set as the Daubechies wavelet, and the decomposition level was three layers. By extracting abrupt change points in the high-frequency subband and combining this with spatial location information, the coordinates of boundary points were determined. It was assumed that the analysis results showed that the boundary points were concentrated within the range of the 300th to 350th sampling points of the signal sequence, corresponding to the actual geological location within the region of 116.5°E to 116.6°E longitude. The entire process was completed automatically by the algorithm. The calculation of the energy ratio and the comparison with a threshold formed a logical chain. If the energy ratio did not exceed the threshold, it could be linked to the deep structure analysis module, automatically switching the analysis target to ensure system continuity. The extraction of boundary information further verified the dominance of shallow details, consistent with the aforementioned judgment results.

[0033] Step S102: Wavelet transform is used to decompose the boundary information of shallow geological bodies at multiple scales to obtain a multi-scale wave coefficient sequence. This process can be divided into six steps, S1021-S1026: Step S1021: Wavelet transform is used to process the boundary data of shallow geology. The signal components at different levels are separated by multi-scale decomposition to obtain a preliminary wave coefficient sequence.

[0034] Step S1022: For the preliminary wave coefficient sequence, perform scale analysis to separate the main signal features at each scale and determine the distribution of geological body information at different scales.

[0035] Step S1023: Obtain boundary feature data from the distribution of geological body information at different scales. By comparing the signal intensity at each scale, determine the significance of the boundary features. If the signal intensity at a certain scale exceeds a preset threshold, mark it as a key boundary feature.

[0036] Step S1024: Based on the marked key boundary features, the wave coefficient sequence is reorganized to generate optimized sequence data, resulting in a feature sequence that better matches the distribution of geological bodies.

[0037] Step S1025: By extracting information from the optimized feature sequence, the core data related to shallow geology is separated, and the final boundary feature description is determined.

[0038] Step S1026: Using the final boundary feature description and combining it with the pre-established geological body information database, perform data matching processing to obtain analysis results consistent with the actual geological body distribution.

[0039] Specifically, regarding the technique of using wavelet transform to decompose shallow geological body boundary information at multiple scales and obtain multi-scale wave coefficient sequences, the implementation method can be summarized as follows: First, assume we have collected a one-dimensional signal data segment of a shallow geological body boundary. The signal length is 1024 sampling points, the sampling frequency is 100Hz, and the signal value range is between -5.0 and 5.0, containing the abrupt change characteristics of the geological body boundary. The initial step is to preprocess the signal, using mean filtering to remove noise. The filter window size is set to 5 points, and the average value of each point is calculated to smooth the data and ensure the accuracy of subsequent decomposition. Next, the Daubechies wavelet (db4) was selected as the basis function, and the signal was decomposed into five levels of multi-scale components using Discrete Wavelet Transform (DWT). Each level of decomposition divided the signal into low-frequency approximation components and high-frequency detail components. For example, the first level yielded 512 approximation coefficients and 512 detail coefficients. The second level further decomposed the approximation coefficients from the first level, yielding 256 approximation coefficients and 256 detail coefficients, and so on up to the fifth level, ultimately forming a multi-scale wavelet coefficient sequence. After decomposition, the characteristics of the coefficients at each level were analyzed. It was found that the high-frequency detail coefficients in the first and second levels reflected the high-frequency characteristics of boundary abrupt changes, with numerical fluctuations ranging from -2.5 to 2.5, while the low-frequency approximation coefficients in the fifth level reflected the overall trend, with numerical values ​​ranging from -1.0 to 1.0. Through statistical analysis of these coefficients, the energy proportion of each level was calculated. For example, the energy of the detail coefficients in the first level accounted for 30% of the total energy, and the energy of the approximation coefficients in the fifth level accounted for 15% of the total energy, thus indicating that boundary information is mainly concentrated in the high-frequency part. Finally, these coefficient sequences are stored in matrix form with 5 rows and 1024 columns for subsequent geological body boundary feature extraction and reconstruction.

[0040] Step S103: The peak features of the multi-scale wave coefficient sequence are extracted using the threshold filtering method. The peak features are then matched with a pre-established fault pattern library using a support vector machine classifier to determine the location of shallow fault folds. This process can be divided into six steps, S1031-S1036: Step S1031: Acquire multi-scale wave data. The original signal is decomposed to generate corresponding coefficient sequence data. A hierarchical processing method is used to decompose the signal into waveform components of different scales, obtaining preliminary sequence results.

[0041] Step S1032: Extract peak feature data from the preliminary sequence results. Using a preset threshold filtering method, if a sequence value exceeds the threshold, it is marked as a peak point, thus determining the peak feature set.

[0042] Step S1033: Apply a support vector machine classifier to perform pattern matching processing using the peak feature set. For each peak point in the feature set, calculate its matching degree with the pre-established tomographic pattern library to obtain the classification result data.

[0043] Step S1034: Based on the classification results data, analyze the pattern features related to shallow faults. Using clustering methods, peak points with high matching degrees are classified as fault-related features to determine the distribution range of fault features.

[0044] Step S1035: Locate the specific position of shallow fault folds by analyzing the distribution range of fault features. Using spatial mapping technology, the feature distribution range is mapped to the actual stratigraphic coordinates to obtain the fold location information.

[0045] Step S1036: Generate structured fault detection data based on the fold location information. Using data integration techniques, the location information is correlated with the feature distribution range to determine the final fault localization result.

[0046] Specifically, the technique of extracting peak features from multi-scale wavelet coefficient sequences and using a support vector machine classifier for pattern matching to determine the location of shallow fault folds can be implemented using the following method. First, assume we obtain a set of coefficient sequences after multi-scale wavelet decomposition from seismic exploration data. The data length is 1000 sampling points, the sampling frequency is 100Hz, and the time window is 10 seconds. Through wavelet transform (using the Daubechies wavelet basis db4), the signal is decomposed into 5 scale layers, obtaining coefficient sequences for each layer. For each layer of coefficients, local peak points are calculated, and a peak detection threshold is set to twice the standard deviation of the sequence. For example, if the standard deviation of a certain layer of coefficients is 3.5, then the threshold is 7.0. Only points with an absolute value greater than 7.0 are retained as candidate peak points, and their positions and amplitudes are recorded. For instance, in the third layer of coefficients, a peak point is detected at sampling point 200 with an amplitude of 8.2. Next, features are extracted from the peak points at all scales to construct feature vectors, including peak amplitude, location, and temporal distance between adjacent points, resulting in a feature matrix of dimension 50×3 (50 peak points, 3 features per point). Subsequently, the feature matrix is ​​input into a support vector machine classifier using a radial basis function (RBF) kernel with a kernel parameter gamma of 0.1 and a regularization parameter C of 1.0. The model is trained using a training dataset (containing labeled data of known fault fold locations, such as a fault location near sampling point 300) to optimize the classification boundary. The classifier outputs the probability of each peak point belonging to its class; points with a probability greater than 0.8 are identified as potential fault fold-related points. For example, sampling point 200 has a probability of 0.85 and is labeled as a potential point. Finally, potential points are clustered using a spatial clustering algorithm (such as DBSCAN, with a neighborhood radius of 10 sampling points and a minimum number of points of 3) to determine the location range of fault folds. For example, the clustering results show that sampling points 190 to 210 form a fault fold region. The rationality of this is verified by comparing it with prior geological knowledge, and the error range is controlled within ±5 sampling points.

[0047] Step S104: Based on the location data of shallow faults and folds and the characteristic information of low-frequency signal components, an inversion algorithm is used to reconstruct the deep structure model and obtain the distribution map of deep geological bodies. By acquiring data on shallow fault and fold locations using automated scanning technology, an initial geological feature dataset was obtained to determine the distribution range of shallow geological morphology. Based on this initial dataset, and combining it with low-frequency signal components, a pre-established signal processing module was used to extract low-frequency feature information related to deep structures, obtaining key signal data for deep structures. For this key signal data, an inversion algorithm was applied to process and generate a preliminary deep structure model, determining the spatial distribution characteristics of geological bodies within the model. If the spatial distribution characteristics of the preliminary deep structure model did not match a preset threshold, signal enhancement technology was used to perform secondary processing on the key signal data, resulting in an optimized signal dataset. Based on the optimized signal dataset, the inversion algorithm was reapplied to update the deep structure model, determining the morphology and distribution patterns of deep geological bodies. Using the updated deep structure model, combined with geological distribution mapping technology, a final distribution map of deep geological bodies was generated, providing a comprehensive representation of the geological features.

[0048] Specifically, in acquiring the locations of shallow fault folds and reconstructing the distribution maps of deep geological bodies, the locations of shallow fault folds were first extracted using high-resolution seismic exploration data and time-frequency analysis methods. Specifically, wavelet transform was employed to decompose the seismic signal into different frequency components, with a frequency range of 10-50Hz, to extract shallow features. The depth error of the calculated fault location was controlled within ±5 meters. The fault line strike was analyzed using signal energy distribution maps, yielding a data model with a fault dip angle of approximately 30 degrees. Subsequently, low-frequency components were fused into the shallow data. Hilbert transform was used to extract low-frequency signals, with a frequency range of 2-8Hz. A weighted superposition algorithm was then used to fuse the low-frequency signals with the shallow data, with a weighting coefficient of 0.6 to ensure signal smoothness. Analysis showed that the signal-to-noise ratio of the fused signal improved to 15dB, verifying the data integrity. Next, based on the fused data, a constrained inversion algorithm was used to reconstruct the deep structural model. The objective function was optimized using the least squares method, with 100 iterations and a convergence error controlled within 0.01. The velocity model of the deep geological body was obtained through inversion calculation, and the velocity variation range was analyzed to be between 2000-3500 m / s, inferring the distribution characteristics of deep rock strata. Finally, based on the inversion results, a distribution map of the deep geological body was generated. Using 3D visualization technology, the velocity model was converted into a geological body distribution image, with a grid resolution of 10 m × 10 m. The clarity of the geological body boundaries was analyzed, and the accuracy of the distribution map was verified by combining stratigraphic thickness data (average thickness approximately 200 m).

[0049] Step S105: Based on the overlap between the distribution map of deep geological bodies and the location of shallow faults and folds, determine the high-risk geological hazard area and obtain the hazard location coordinates; at the same time, determine the propagation interference area through the hazard location coordinates. Data on the distribution of deep geological bodies and the location of shallow faults and folds were acquired. Data preprocessing was performed to unify the format and calibrate the coordinate system of both datasets, resulting in a standardized geological distribution dataset. For this standardized dataset, an overlay analysis method was used to spatially overlap and compare the deep geological body distribution data with the shallow fault and fold location data to determine the distribution range of overlapping areas. Based on the distribution range of overlapping areas, a preset threshold was used. If the proportion of overlapping areas exceeded the preset threshold, these areas were identified as potentially high-risk areas, and preliminary identification of these areas was obtained. Using this preliminary identification and historical data on geological hazards, a support vector machine algorithm was employed for risk classification to determine the specific category and hazard level of the high-risk areas. After obtaining the specific categories and hazard levels of the high-risk areas, coordinates were extracted for each category to determine the precise distribution of hazard location coordinates. Based on the precise distribution of hazard location coordinates, spatial mapping was performed using the distribution map data to obtain a detailed distribution layer of the high-risk areas. This detailed distribution layer of high-risk areas was then overlaid with shallow fault and deep geological body data for multi-level spatial analysis to determine the final key monitoring areas for geological hazards.

[0050] Specifically, when performing overlay analysis of deep geological body distribution maps and shallow fault / fold locations, the distribution data of deep geological bodies and the location data of shallow faults / folds are first obtained using Geographic Information System (GIS) software. It is assumed that the deep geological body distribution map is stored in a grid format, with each grid cell measuring 10 meters × 10 meters. The data includes the depth values ​​of the geological bodies, ranging from -500 meters to -2000 meters. The shallow fault / fold location data is stored as vector lines, containing fault line length and dip information, with lengths ranging from 100 meters to 5000 meters. Next, the two sets of data are spatially overlaid using a spatial intersection algorithm to calculate the overlapping area of ​​the two layers in the same coordinate system. Specifically, a projection transformation is used to unify the data to the WGS84 coordinate system. By calculating the intersection area of ​​each grid cell with the fault line vector, the percentage of overlap is obtained. For example, if the total area of ​​a certain region's grid cells is 100 square meters and the intersection area with the fault line is 60 square meters, then the overlap percentage is 60%. Subsequently, a preset threshold of 50% is set. If the overlap rate of a certain area exceeds 50%, it is marked as a potential risk area. For example, the area mentioned above, with an overlap of 60%, exceeds the threshold of 50% and is therefore marked. Further, a risk assessment is performed on the marked areas. Combining geological depth values ​​and fault dip angle data, a weighted scoring algorithm is used. Assuming a depth weight of 0.6 and a dip angle weight of 0.4, a depth of -1500 meters corresponds to a risk score of 80, and a dip angle of 30 degrees corresponds to a risk score of 70. Therefore, the total risk score for this area is 80 × 0.6 + 70 × 0.4 = 76 points. If the risk score exceeds the preset value of 60 points, it is judged as a high-risk geological hazard area. Finally, the coordinates of the center point of this area are extracted as the hazard location coordinates. Assuming the calculated coordinates are (longitude 116.35, latitude 39.85), these coordinates are stored in the database, and a risk report is generated, containing the coordinates, the risk score of 76, and relevant geological parameters.

[0051] Initial coordinate data is obtained by locating potential hazards. Outliers are removed using data cleaning techniques to obtain a processed location dataset. Based on this dataset, a three-dimensional mesh structure is constructed, and the space is divided into multiple units using mesh generation techniques to determine the spatial distribution framework. For this framework, seismic wave propagation parameters are loaded, and numerical simulation using the finite difference method is performed to obtain preliminary wavefield propagation results. Propagation trajectory data is extracted from these preliminary results. If significant deviations occur in the trajectory data within certain mesh units, these areas are marked as potential interference regions, resulting in a preliminary interference distribution. Using this preliminary distribution and simulation analysis techniques, the potential interference regions are further verified to determine their boundaries and obtain precise location results. Based on the precise location results, spatial distribution data of the interference regions is generated, and visualization techniques are used to output the distribution of interference regions within the three-dimensional mesh, thus determining the final analysis results.

[0052] Specifically, in the process of generating a three-dimensional visualization grid based on the hazard location coordinates and simulating the seismic wave propagation path to determine the propagation interference zone, the process begins by using the hazard location coordinate data. For example, assuming the hazard point is located at spatial coordinates (100.5, 200.3, 50.2) meters, a computer automatically generates a three-dimensional mesh model. The mesh unit size is set to 1 meter × 1 meter × 1 meter, covering a cubic area of ​​500 meters × 500 meters around the hazard point, generating a total of 125,000,000 mesh units. The algorithm automatically marks the hazard point as a core node, and a three-dimensional visualization algorithm (such as volume data rendering) is used to construct a visual representation of the mesh in the computer. The relative position of the hazard point in the mesh is analyzed to determine whether it is located at the center or boundary of the mesh. If it is located at the boundary, the mesh range is automatically expanded to 600 meters to ensure that the hazard point is centered. Next, the finite difference method was used to simulate the seismic wave propagation path. Assuming an initial seismic wave velocity of 3000 m / s and a medium density of 2500 kg / m³, a discretized model was constructed based on the wave equation, with a time step of 0.001 seconds and a spatial step of 1 meter. The stress and velocity components of the wave field in each grid cell were iteratively calculated to analyze the energy changes as the wave passed through the hazard point during propagation. The wavefront arrival time and amplitude attenuation were recorded. For example, an amplitude attenuation of 30% was found within 100 meters of the hazard point, indicating significant interference. Finally, the propagation interference zone was determined. Based on the simulation results, grid areas with amplitude attenuation exceeding 20% ​​were automatically extracted, and the calculated volume of the interference zone was approximately 5236 cubic meters. The boundary of the interference zone was further analyzed in conjunction with wave velocity changes (decreasing to 2800 m / s), generating a three-dimensional distribution map of the interference zone. The overlap between the hazard point and the interference zone was automatically marked, indicating an 85% overlap, providing data support for subsequent risk assessment. The above process is fully automated through computer algorithms. Mesh generation relies on spatial partitioning algorithms, wave propagation simulation relies on numerical calculations, and interference zone determination relies on data threshold extraction, forming a process from coordinate input to interference zone output.

[0053] Step S106: By adjusting the weights of the frequency bands in the propagation interference zone of the seismic wave signal, an enhanced resolution signal is generated; By acquiring seismic wave signal data, and targeting the signal characteristics within the interference area, a pre-defined signal separation method is used to obtain an initial set of signal components. Based on the initial set of signal components, adaptive filtering technology is used to adjust the weights according to the frequency component distribution, resulting in adjusted frequency component data. For the adjusted frequency component data, the signal fluctuation characteristics within the interference area are analyzed to determine the interference-related frequency range. If the interference-related frequency range exceeds a preset threshold, the frequency components within that range undergo secondary filtering to obtain filtered signal components. Based on the filtered signal components and the time-domain characteristics of the seismic wave signal, the presence of residual interference is determined, resulting in an interference determination result. Using the interference determination result, a signal superposition method is used to enhance the waveform for the residual interference portion, obtaining the final enhanced resolution signal output. If interference characteristics are still detected in the final enhanced resolution signal output, local frequency adjustment is performed on the signal output to determine the final optimized signal data.

[0054] Specifically, the technique of adjusting the frequency component weights in the propagation interference zone and using an adaptive filter to optimize the seismic wave signal to obtain a signal with enhanced resolution can be implemented through the following specific methods. First, regarding the adjustment of the frequency component weights in the propagation interference zone, assuming the received seismic wave signal contains multiple frequency components, the signal is decomposed into frequency domain data using Fourier transform, obtaining a spectrum ranging from 0.5Hz to 50Hz. Analysis reveals that the interference zone is mainly concentrated in the 10Hz to 20Hz frequency band, accounting for as much as 60% of the energy. Therefore, a frequency weight adjustment algorithm is designed, setting the weight coefficient for the 10Hz to 20Hz frequency band to 0.3, while setting the weight coefficients for other frequency bands such as 0.5Hz to 10Hz and 20Hz to 50Hz to 0.8. Through weighted processing, the influence of the interference frequency band is reduced. After resynthesizing the signal, the interference energy proportion decreases to 25%, and the signal-to-noise ratio improves by approximately 3dB. Subsequently, to optimize the seismic wave signal, an adaptive filter, such as the Least Mean Square Error (LMS) algorithm, was employed. An initial step size of 0.01 and a filter order of 32 were set. The filter coefficients were iteratively updated to minimize the mean square error between the output signal and the desired signal. After 1000 iterations, the error converged to 0.002, preserving signal details and improving resolution by approximately 15%. Finally, to obtain a signal with enhanced resolution, the filtered signal underwent wavelet transform using the Daubechies wavelet basis. A five-layer decomposition was performed, extracting high-frequency detail components and amplifying their amplitude by approximately 1.2 times, while suppressing low-frequency noise components and reducing their amplitude by 0.5 times. After reconstructing the signal, the resolution was further improved to 20%, and the signal-to-noise ratio reached over 5 dB.

[0055] Step S107: Reconstruct the overall geological model based on the enhanced resolution signal, determine the interaction between geological hazards and the route corridor in the model, and obtain the final risk prediction result.

[0056] This process can be divided into seven specific steps, S1071-S1077: Step S1071: Obtain the resolution signal from the collected data, and process the signal using signal enhancement technology to obtain enhanced signal data.

[0057] Step S1072: Based on the enhanced signal data, reconstruct the geological model using a pre-established geological modeling method to determine the structural distribution of the geological model.

[0058] Step S1073: Analyze the distribution of geological hazards in the geological model's structure and obtain the specific location information of the hazards.

[0059] Step S1074: Extract the interaction impact data related to the line corridor from the specific location information of the hidden danger distribution, determine whether the corridor location is affected by the hidden danger distribution, and if so, record the affected area range.

[0060] Step S1075: Based on the affected area and the interaction between geological hazards and corridor location, the risk level is classified using the support vector machine algorithm to obtain preliminary risk assessment results.

[0061] Step S1076: Based on the preliminary results of the risk assessment and the structural distribution data from the model analysis, determine the final risk assessment conclusion.

[0062] Step S1077: Based on the final risk assessment conclusion, generate the corresponding risk distribution layer and obtain visualized risk distribution information.

[0063] Specifically, in constructing an enhanced resolution signal to reconstruct the overall geological model and assess the interaction between geological hazards and the railway corridor, the signal was first enhanced using high-resolution seismic exploration data and ground subsidence monitoring data. Wavelet transform algorithms were then used to denoise the original signal, with a frequency threshold of 5.0 Hz set to filter out low-frequency interference. Subsequently, an inversion algorithm was used to reconstruct the three-dimensional geological model, increasing the resolution to 0.5 meters per pixel, ensuring clear visibility of details such as rock layer boundaries and fault zones. Next, geological hazard features were extracted based on the model. For example, a fault zone with a width of 2.0 meters, a dip angle of 45 degrees, and a depth ranging from 10.0 to 15.0 meters underground was identified. Finite element analysis was used to calculate its impact on surface subsidence, yielding a maximum subsidence of 0.03 meters per year. Then, the planning data of the route corridor was imported into the model. Assuming the corridor width is 50.0 meters and the length is 10.0 kilometers, the spatial overlay analysis algorithm was used to determine its interaction with the fault zone. It was found that a 3.0-kilometer section of the corridor overlaps with the fault zone, and the geological stress concentration factor of the overlapping area was calculated to be 1.5, indicating a high risk. Finally, based on the distribution of geological hazards and the results of the interaction analysis, the risk assessment matrix method was adopted, and a risk level threshold of 3.0 was set. The risk value of the overlapping area was calculated to be 4.2, exceeding the threshold. A final risk assessment report was generated, which indicated that reinforcement measures should be prioritized for this section and recommended adjusting the corridor route to avoid the fault zone within a 2.0-kilometer range.

[0064] Example 3: This embodiment discloses an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements a method for detecting and predicting adverse geological features in a power transmission and distribution line corridor as described in any embodiment of the present invention.

[0065] Example 4: A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements a method for detecting and predicting adverse geological features in a power transmission and distribution line corridor as described in any embodiment of the present invention.

[0066] The above description is merely an embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the content of the present invention's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A method for detecting and predicting risks of adverse geological features in power transmission and distribution line corridors, characterized in that, The specific steps include: Seismic wave signal data were collected from the power transmission and distribution line corridor area. High-frequency and low-frequency signal components were separated using time-frequency analysis to obtain frequency distribution characteristics. Based on the frequency distribution characteristics of the seismic wave signals, the boundary information of shallow geological bodies was obtained. Wavelet transform was used to decompose the boundary information of shallow geological bodies at multiple scales to obtain a multi-scale wave coefficient sequence. The peak features of the multi-scale wave coefficient sequence were extracted using a threshold filtering method. The peak features were then matched with a pre-established fault pattern library using a support vector machine classifier to determine the location of shallow fault folds. Based on the location data of shallow faults and folds and the characteristic information of low-frequency signal components, an inversion algorithm is used to reconstruct the deep structural model and obtain the distribution map of deep geological bodies. High-risk geological hazard areas are identified by the overlap between the distribution map of deep geological bodies and the location of shallow faults and folds, and the coordinates of the hazards are obtained. At the same time, the areas of propagation interference are determined by the coordinates of the hazards. An enhanced resolution signal is generated by adjusting the weights of the frequency bands in the interference zone of the seismic wave signal. The overall geological model is reconstructed based on the enhanced resolution signal, and the interaction between geological hazards and the route corridor in the model is determined to obtain the final risk prediction results.

2. The method for detecting and predicting risks of adverse geological features in power transmission and distribution line corridors according to claim 1, characterized in that, The specific steps for obtaining shallow geological body boundary information based on the frequency distribution characteristics of seismic wave signals are as follows: Calculate the energy ratio of high-frequency signal components to low-frequency signal components based on frequency distribution characteristics; If the energy ratio is greater than the preset threshold, it is determined to be a shallow detail-dominated region, and the boundary information of the shallow geological body in that region is extracted.

3. The method for detecting and predicting risks of adverse geological features in power transmission and distribution line corridors according to claim 1, characterized in that, The specific steps for using wavelet transform to perform multi-scale decomposition of shallow geological body boundary information and obtain multi-scale wave coefficient sequences are as follows: Wavelet transform was used to decompose the boundary information of shallow geological bodies at multiple scales to obtain a preliminary wave coefficient sequence. Key boundary features are marked based on signal intensity at different scales; The preliminary wave coefficient sequence is reorganized based on key boundary features to obtain an optimized multi-scale wave coefficient sequence.

4. The method for detecting and predicting risks of adverse geological features in power transmission and distribution line corridors according to claim 1, characterized in that, The step of matching peak features with a pre-established fault pattern library using a support vector machine classifier to determine the location of shallow fault folds specifically involves: The peak features are input into a support vector machine classifier and matched with a pre-set tomographic pattern library to obtain the classification result. Then, the location of shallow fault folds is determined based on the classification results.

5. The method for detecting and predicting risks of adverse geological features in power transmission and distribution line corridors according to claim 1, characterized in that, The specific steps for reconstructing a deep structural model and obtaining a deep geological body distribution map based on shallow fault and fold location data and low-frequency signal component feature information using an inversion algorithm are as follows: Extract relevant low-frequency signal components based on the location of shallow fault folds; Extracting key signal data related to deep structures from low-frequency signal components; Based on key signal data, a deep structural model is generated through an inversion algorithm, and a distribution map of deep geological bodies is obtained by combining it with geological mapping technology.

6. The method for detecting and predicting risks of adverse geological features in power transmission and distribution line corridors according to claim 1, characterized in that, The specific steps for determining high-risk geological hazard areas based on the overlap between the distribution map of deep geological bodies and the location of shallow faults and folds are as follows: The distribution map of deep geological bodies is spatially overlaid with the location of shallow faults and folds, and the percentage of overlapping areas is calculated. If the percentage of overlapping areas is greater than a preset threshold, it is determined to be a high-risk geological hazard area.

7. The method for detecting and predicting risks of adverse geological features in power transmission and distribution line corridors according to claim 1, characterized in that, The specific steps for determining the propagation interference zone using the hazard location coordinates are as follows: A three-dimensional mesh structure is constructed based on the coordinates of the hazard location, and then seismic wave propagation parameters are loaded based on the three-dimensional mesh structure; Numerical simulations were performed using the finite difference method to extract the propagation trajectory of seismic waves. Grid cells whose deviation from the seismic wave propagation trajectory exceeds a preset threshold are identified as propagation interference zones.

8. The method for detecting and predicting risks of adverse geological features in power transmission and distribution line corridors according to claim 6, characterized in that, The step of generating an enhanced resolution signal by adjusting the weights of the frequency bands in the propagation interference zone of the seismic wave signal is as follows: For the signal frequency range of the propagation interference zone, the frequency ranges of the propagation interference zone and the non-propagation interference zone are distinguished from the frequency range of the seismic wave signal to obtain the initial set of seismic wave signal frequency components. Set the weights for each frequency range of the initial seismic wave signal frequency component set, and resynthesize and optimize the seismic wave signal through weighted processing; An adaptive filter is used to perform secondary filtering on the optimized seismic wave signal to obtain the filtered signal components. The filtered signal components are then enhanced to obtain a signal with enhanced resolution.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements a method for detecting and predicting adverse geological features in power transmission and distribution line corridors as described in any one of claims 1 to 8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by the processor, the program implements a method for detecting and predicting adverse geological features in power transmission and distribution line corridors as described in any one of claims 1 to 8.

Images