Rapid prediction method and system for small-scale collapse column of coal mine

Abstract

The invention discloses a rapid prediction method and system for a small-scale collapse column of a coal mine. Determining a seismic event corresponding to the target coal seam, and generating a target horizon; analyzing the three-dimensional seismic data in a target horizon and a neighborhood range thereof, and establishing a comprehensive abnormal response; identifying a space boundary zone transited from the stable zone to the disturbance zone, and constructing a closed or semi-closed boundary ring zone; selecting a plurality of seismic sections, extracting candidate central positions corresponding to the space of the comprehensive abnormal response in each seismic section, and determining the position with the highest consistency as a central control point of the collapse column; constructing a plurality of seismic sections, identifying an abnormal response position near the central control point, and obtaining a plurality of candidate boundary control points; forming an effective boundary point set; generating a plane boundary contour of the collapse column, and obtaining an influence range of the collapse column on the target coal seam; and checking and verifying the spatial relationship between the plane boundary contour and the target coal seam. And the interpretation efficiency is remarkably improved while the prediction precision is ensured.

Description

Technical Field

[0001] This invention relates to the field of fine identification technology of small-scale collapse columns in coal mines, and more specifically to a rapid prediction method and system for small-scale collapse columns in coal mines. Background Technology

[0002] Currently, 3D seismic exploration technology, with its advantages of wide coverage, high spatial resolution, and rich information, is gradually becoming an important technical means for fine-grained detection of coal mine structures. Based on 3D seismic data, existing studies have analyzed the seismic response characteristics of collapse columns through methods such as seismic time profile analysis and multi-attribute interpretation, achieving certain results in the identification of medium- and large-scale collapse columns (major axis diameter greater than 50m).

[0003] However, traditional interpretation methods have significant limitations for small-scale collapse columns. On the one hand, the anomalous responses of small-scale collapse columns are not prominent enough on seismic time profiles and attribute planes, making it difficult to accurately characterize their true boundaries. On the other hand, when multiple collapse columns are adjacent to each other on a plane, the seismic anomalous responses superimpose, increasing the uncertainty of the interpretation results. Furthermore, the seismic response characteristics of anomalous geological phenomena such as small grabens, small synclines, and local dip abrupt changes are similar to those of small-scale collapse columns, leading to a high risk of misjudgment and affecting the reliability of the interpretation results.

[0004] Therefore, how to solve the technical problems of slow prediction speed and low accuracy of small-scale collapse columns in existing technologies, and provide a guarantee for safe and efficient mining in coal mines, is an urgent problem that needs to be solved by those skilled in the art. Summary of the Invention

[0005] In view of the above problems, the present invention is proposed to provide a rapid prediction method and system for small-scale collapse columns in coal mines that overcomes or at least partially solves the above problems.

[0006] To achieve the above objectives, the present invention adopts the following technical solution:

[0007] In a first aspect, embodiments of the present invention provide a rapid prediction method for small-scale collapse columns in coal mines, specifically including the following steps: S1. Based on the well-seismic calibration results, determine the seismic phase axis corresponding to the target coal seam, and use the automatic recursive method of spatial small-area elements to interpret the stratigraphic position and generate the target stratigraphic position. S2. Analyze the three-dimensional seismic data within the target layer and its neighborhood to establish a comprehensive anomaly response. The comprehensive anomaly response characterizes the local disturbance of the reflecting structure, wavefield consistency and phase and structural joint stability changes, and delineates the seismic response characteristics of potential small-scale collapse columns. S3. Based on the spatial distribution and continuity characteristics of the comprehensive anomaly response on a planar scale, identify the spatial boundary zone transitioning from the stable region to the disturbance region, and construct the spatial boundary zone into a closed or semi-closed boundary ring. S4. Within the constraint range of the closed or semi-closed boundary ring, select multiple seismic profiles, extract the candidate center positions in each seismic profile that correspond to the spatial relationship of the integrated anomaly response, and analyze the spatial consistency of the candidate center positions in different orientation profiles to determine the position with the highest consistency as the center control point of the collapse column. S5. Using the central control point as a reference, construct multiple seismic profiles at preset azimuth intervals within the target layer plane, identify anomalous response locations near the central control point in each seismic profile, and obtain multiple candidate boundary control points. S6. Map the candidate boundary control points to the target layer plane in a unified manner, and combine the spatial consistency characteristics of the comprehensive anomaly response to eliminate or correct candidate boundary control points that do not meet the preset spatial constraints, so as to form an effective boundary point set. S7. Based on the effective boundary point set, generate the planar boundary profile of the collapse column in the target stratum plane through an adaptive boundary fitting method, and obtain the influence range of the collapse column on the target coal seam. S8. Verify the spatial relationship between the planar boundary contour and the target coal seam using three-dimensional visualization.

[0008] Furthermore, the target layer in step S1 is obtained as follows: Using a manually calibrated sample profile as an initial constraint, the sample profile is divided into several three-dimensional micro-elements. The target layer is recursively generated in three-dimensional space based on the waveform feature comparison method, while allowing manual correction of the recursion process in real time.

[0009] Furthermore, the profile anomaly mode of the comprehensive anomaly response described in step S2 includes: local disturbances and multi-layer "concave" deformation features of the in-phase axis, multiple accompanying phase features above and below the in-phase axis in the collapse column development area, features of local absence or reduced continuity of the in-phase axis, features of weakened or disappeared delayed diffraction wave response, features of decreased phase continuity of reflected wave but not complete disorder, features of the in-phase axis transitioning from a single-axis response to a dual-axis or multi-axis response and rapidly recovering, and features of phase tailing at the boundary of collapse column development.

[0010] Furthermore, the comprehensive abnormal response described in step S2 is a combination of one or more of the following characteristics: accompanying phase response appears above and below the phase axis, while the phase continuity decreases but does not become completely disordered, and the phase axis changes from a single-axis response to a biaxial or multiaxial response. This is used as a criterion for distinguishing small-scale collapse columns from faults, folds or coal seam scour zones.

[0011] Furthermore, in step S3, the closed or semi-closed boundary ring is formed by identifying the gradient change characteristics of the integrated anomaly response on a planar scale, wherein the gradient change characteristics include the abrupt change characteristics of the anomaly from the stable region to the disturbed region.

[0012] Furthermore, in step S3, the closed or semi-closed boundary ring is determined by the spatial superposition relationship of multiple types of integrated anomalous responses, and the multiple types of anomalous responses include at least reflection structure disturbance anomalies and wave field consistency change anomalies.

[0013] Furthermore, in step S4, the spatial consistency analysis of the candidate center location includes comparing the spatial concentration or recurrence frequency of abnormal locations in different orientation profiles.

[0014] Furthermore, in step S6, the constraint of spatial consistency includes: the distribution of candidate boundary control points on the target stratigraphic plane must correspond to the spatial morphology of the closed or semi-closed boundary ring.

[0015] Furthermore, in step S7, the adaptive boundary fitting method smooths the boundary contour based on the distribution density change of the effective boundary point set on the plane.

[0016] Secondly, embodiments of the present invention provide a rapid prediction system for small-scale collapse columns in coal mines, comprising: A stratigraphic construction module, which generates the target coal seam stratigraphic level based on well seismic calibration results and combined with an automatic recursive method using small spatial elements; An anomaly response analysis module analyzes three-dimensional seismic data within the target layer and its neighborhood to extract a comprehensive anomaly response for characterizing potential small-scale collapse columns. An anomaly concentration area identification module identifies the spatial boundary zone of the integrated anomaly response on a planar scale and constructs a closed or semi-closed boundary ring for constraining the influence range of potential collapse columns. The central control point determination module performs spatial consistency analysis on candidate center locations in multiple seismic profiles within the boundary ring constraint range to determine the central control point of the collapse column. A multi-directional interpretation module, which constructs a multi-directional seismic profile within the target horizon plane and obtains candidate boundary control points based on the central control point; The joint constraint verification module maps candidate boundary control points to the target layer plane and performs screening and correction based on the spatial consistency characteristics of the comprehensive anomaly response. A boundary generation module generates the planar boundary profile of the collapsed column based on an effective set of boundary points using an adaptive boundary fitting method. The three-dimensional verification module verifies the spatial relationship between the planar boundary contour and the target coal seam through three-dimensional visualization.

[0017] As can be seen from the above technical solution, compared with the prior art, the present invention discloses a rapid prediction method and system for small-scale collapse columns in coal mines, which has the following beneficial effects: This invention provides a rapid identification method and system for small-scale collapse columns, enabling accurate interpretation of these columns. This technical solution effectively overcomes the problems of weak seismic response, low boundary accuracy, and strong subjectivity in interpretation associated with small-scale collapse columns. While ensuring prediction accuracy, it significantly improves interpretation efficiency and is suitable for fine detection of small-scale collapse columns in coal mines under complex geological conditions. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of the present 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 embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0019] Figure 1 This is a flowchart of a rapid prediction method for small-scale collapse columns in coal mines provided in an embodiment of the present invention. Figures 2a-2e This is a schematic diagram of a typical integrated anomaly response provided in an embodiment of the present invention; Figure 3 This is a structural diagram of the rapid prediction system for small-scale collapse columns in coal mines provided in an embodiment of the present invention. Detailed Implementation

[0020] 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.

[0021] This invention discloses a rapid prediction method for small-scale collapse columns in coal mines, such as... Figure 1 As shown, it includes the following steps: S1. Based on the well-seismic calibration results, determine the seismic phase axis corresponding to the target coal seam, and use the automatic recursive method of spatial small-area elements to interpret the stratigraphic position and generate the target stratigraphic position. S2. Analyze the three-dimensional seismic data within the target layer and its neighborhood to establish a comprehensive anomaly response. The comprehensive anomaly response characterizes the local disturbance of the reflecting structure, wavefield consistency and phase and structural joint stability changes, and delineates the seismic response characteristics of potential small-scale collapse columns. S3. Based on the spatial distribution and continuity characteristics of the comprehensive anomaly response on a planar scale, identify the spatial boundary zone transitioning from the stable region to the disturbance region, and construct the spatial boundary zone into a closed or semi-closed boundary ring. S4. Within the constraint range of the closed or semi-closed boundary ring, select multiple seismic profiles, extract the candidate center positions in each seismic profile that correspond to the spatial relationship of the integrated anomaly response, and analyze the spatial consistency of the candidate center positions in different orientation profiles to determine the position with the highest consistency as the center control point of the collapse column. S5. Using the central control point as a reference, construct multiple seismic profiles at preset azimuth intervals within the target layer plane, identify anomalous response locations near the central control point in each seismic profile, and obtain multiple candidate boundary control points. S6. Map the candidate boundary control points to the target layer plane in a unified manner, and combine the spatial consistency characteristics of the comprehensive anomaly response to eliminate or correct candidate boundary control points that do not meet the preset spatial constraints, so as to form an effective boundary point set. S7. Based on the effective boundary point set, generate the planar boundary profile of the collapse column in the target stratum plane through an adaptive boundary fitting method, and obtain the influence range of the collapse column on the target coal seam. S8. Verify the spatial relationship between the planar boundary contour and the target coal seam using three-dimensional visualization.

[0022] This invention provides a rapid identification method for small-scale collapse columns, enabling accurate interpretation of these columns. This technical solution effectively overcomes the problems of weak seismic response, low boundary identification accuracy, and strong subjectivity in interpretation associated with small-scale collapse columns. While ensuring prediction accuracy, it significantly improves interpretation efficiency and is suitable for fine detection of small-scale collapse columns in coal mines under complex geological conditions.

[0023] The following is a detailed description of each of the above steps: In step S1, the target coal seam horizon is constructed; The phase axis corresponding to the target coal seam is determined based on the well logging and seismic calibration results. In practice, the relationship between well logging data and seismic data can be used to clarify the location of the target coal seam on the seismic profile.

[0024] Based on this, an automatic recursive method using small spatial elements is employed to interpret the stratigraphic positions of the target coal seam. Specifically, interpreters manually calibrate the phase axis of the target coal seam on a sample profile as an initial constraint. Subsequently, the sample profile is divided into several three-dimensional micro-elements, and based on waveform feature comparison, recursive tracing is performed in three-dimensional space to gradually generate a continuous representation of the target coal seam's stratigraphic position. During the recursive process, interpreters are allowed to make real-time corrections to the automatic interpretation results based on geological understanding. The system then uses the corrected results as new constraints to continue the recursive process, thereby achieving rapid and accurate interpretation of the target stratigraphic positions across the entire area.

[0025] In one implementation, a sample profile is first manually interpreted on the target phase axis determined by well-seismic calibration. This interpretation result serves as the constraint and training input for the model. Based on the manually interpreted sample profile, the system automatically collects information such as waveform, phase, planar trend, and attitude of the target strata. The sample profile is then decomposed into several three-dimensional micro-elements according to dip angle. These "micro-elements" are tracked, and the optimal interpretation scheme for the next survey line is automatically generated. If the automatically generated optimal scheme is inconsistent with geological understanding, the interpreter can intervene and correct it in real time during the tracking process. The system updates the interpretation scheme with the corrected results, continues to track the next line, adjusts the stratigraphy along the main survey line direction, and automatically closes the connecting survey line direction, ultimately achieving stratigraphic tracking across the entire area.

[0026] In step S2, a comprehensive anomaly response model is constructed; After the target layer is constructed, the 3D seismic data in the target layer and its neighborhood are analyzed to establish a comprehensive anomaly response model for characterizing potential small-scale collapse columns.

[0027] The comprehensive anomaly response model is used to characterize the local disturbance features of the reflecting structure, changes in wavefield propagation consistency, and changes in the combined stability of phase and structure. Specifically, this comprehensive anomaly response can be obtained by analyzing the geometric changes in the reflected waveform, the changes in the continuity and stability of the phase axis, and the changes in wavefield structural characteristics on the seismic profile, in order to reflect the comprehensive influence of small-scale collapsed columns on the propagation and reflection characteristics of seismic waves.

[0028] The specific process of step S2 is as follows: Step S21: Extraction and preprocessing of target layer neighborhood data; After the target coal seam horizon is constructed, three-dimensional seismic data volumes within a certain depth range are extracted from the target horizon as a reference within a preset time window above and below it. The neighborhood range is used to cover the main influence area of ​​the collapse column that disturbs the target coal seam.

[0029] Necessary preprocessing is performed on the seismic data volume, including but not limited to amplitude equalization, phase correction, or seismic trace consistency adjustment, to reduce the impact of acquisition and processing differences on subsequent anomaly analysis.

[0030] Step S22: Extraction of local perturbation features of the reflection structure; Within the neighborhood of the target horizon, the seismic profile is analyzed channel by channel to extract anomalous information characterizing local geometric changes in the reflecting structure. These local disturbance features of the reflecting structure include, but are not limited to: Local bending or concavity of the same phase shaft; The geometry of the in-phase axes becomes disordered over a short distance; The upper and lower adjacent phase axes exhibit coordinated deformation at the same position.

[0031] The above features are identified to characterize the local structural disturbances caused by the collapse column.

[0032] Step S23: Extraction of wavefield uniformity variation features; Based on the seismic waveform information of the target layer and its neighborhood, the spatial consistency of the seismic wavefield is analyzed to extract anomalous features that characterize changes in wavefield consistency.

[0033] The wavefield uniformity variation characteristics include, but are not limited to: The waveform similarity between adjacent earthquake phase axes decreases; The continuity of the phase axis is weakened, and it changes from regular to discrete. The waveform changes abruptly when transitioning from the stable region to the disturbed region.

[0034] These features are used to reflect the disturbance effect of collapsed columns on the propagation path and wavefield structure of seismic waves.

[0035] Step S24: Extraction of joint stability change features of phase and structure; Based on steps S22 and S23, the phase characteristics of the reflected wave are further analyzed to extract joint anomaly information of phase stability and reflection structure change.

[0036] The combined phase and structure stability variation characteristics include, but are not limited to: Phase continuity decreased but did not become completely disordered; Phase tailing, phase broadening, or phase shift phenomena; Accompanying phase, biaxial or multiaxial phenomena appear near the main phase axis.

[0037] These features are used to characterize the perturbation response of small-scale collapse columns to the coal seam phase axis.

[0038] Step S25: Construct a comprehensive anomaly response model; By comprehensively analyzing the local disturbance characteristics of the aforementioned reflection structure, the wavefield uniformity change characteristics, and the combined stability change characteristics of phase and structure, a comprehensive anomaly response model is formed to characterize the seismic response features of small-scale collapsed columns.

[0039] In the comprehensive anomaly response mode, it is not required that a single anomaly feature appear independently. Instead, the development location and degree of influence of potential small-scale collapse columns are comprehensively identified by the correspondence, superposition, or synergistic change characteristics of multiple anomaly features in spatial location.

[0040] Based on actual geological data, the characteristics of small-scale collapse columns in 3D seismic data are as follows: I. Cross-sectional features include: (1) Local disturbance of the same phase axis, resulting in a "concave" phenomenon. Local disturbances and "concave" phenomena in the phase axis: In areas with small-scale collapse columns, the phase axis of the target coal seam typically exhibits slight local concavity and irregular undulations in its geometric shape. This manifests as increased bending and curvature of the phase axis over a short distance, with the concave area often lacking a stable fracture displacement interface. Unlike the "neat faulting" caused by conventional faults, this type of concavity is more of a local deformation response of a continuous reflection interface: the overall phase axis remains traceable, but local "collapse-like" bending occurs within the collapse column's influence zone.

[0041] In terms of waveform representation, the concave area is often accompanied by the display feature of "thickening" of the phase axis, causing the reflected wave to change from the original relatively sharp single-peak or narrow-band waveform to a wider waveform; at the same time, local amplitude fluctuations, slight phase shifts, and the phenomenon of texture changing from ordered to semi-disordered may appear.

[0042] In terms of longitudinal integration, besides the concavity of the primary reflection phase axis, the secondary phase axes below it often exhibit synchronous concavity or slight bending enhancement at the same plane position, forming a characteristic of "multi-layered co-positional response." This coordinated deformation of the upper and lower wave groups at the same location has strong indicative significance for collapse columns. Furthermore, for small-scale collapse columns, this type of concavity response is not accompanied by obvious interruption or strong displacement of the phase axis; that is, although the phase axis shows local bending and waveform disturbance, it is still generally continuous and traceable. This contrasts with the strong damage, large-segment loss or obvious breakage of the phase axis commonly seen in medium and large-scale collapse columns, and can serve as one of the important discriminative clues for distinguishing between small-scale and medium-to-large-scale collapse columns.

[0043] In this embodiment of the invention, the aforementioned "local disturbance and slight depression" anomalies are incorporated into an important component of the comprehensive anomaly response mode: on the one hand, they are used to quickly locate the influence position of potential collapse columns on multi-directional profiles; on the other hand, they are used to perform spatial correspondence verification with the anomaly boundary zone at the planar scale, thereby providing a stable profile basis for subsequent determination of central control points and screening of boundary control points. By jointly discriminating features such as synchronous depression, widening, and decreased waveform consistency of the coal seam and its upper and lower phase axes, the accuracy of small-scale collapse column identification can be improved.

[0044] (2) Phase trailing In areas with well-developed small-scale collapse columns, a distinct "phase tail" phenomenon is often observed in seismic profiles of coal seam phase axes. For example... Figure 2a As shown, the blue line represents the profile outline of the collapse column, the pink line represents the target layer, and the dashed line indicates the phase axis, which is a typical phase tail response location. This feature is characterized by the fact that the reflected phase axis no longer presents a clear and sharp phase concentration band in the local area, but instead stretches longitudinally after the main reflection, forming a phase tail. The phase continuity of the tail segment is significantly reduced, but the whole still maintains a spatial relationship with the main phase axis.

[0045] Phase tailing is typically not accompanied by significant amplitude enhancement or independent reflecting interfaces. Its formation mechanism is closely related to the complex wave field propagation caused by the interior of the collapsed column and its fracturing zone. Due to the heterogeneity of the collapsed infill material properties, the reduction in wave velocity, and the presence of local multipath propagation effects, reflected waves experience delay superposition and phase broadening when passing through the affected area of ​​the collapsed column, thus forming a tail-like phase response on the time profile. This phenomenon is particularly pronounced at the boundary of the collapsed column and often occurs simultaneously with accompanying phases, biaxial or multiaxial structures.

[0046] Unlike the anomalous response characteristics caused by general structural undulations or slow coal seam thinning, phase tailing exhibits strong locality and obvious asymmetric distribution. Therefore, phase tailing can serve as an important criterion for identifying small-scale collapse columns, especially when amplitude anomalies are not significant or the continuity of reflected waves has not been completely disrupted. It can effectively reveal the disturbances caused by the collapse column, providing a strong basis for the precise identification and boundary constraints of small-scale collapse columns.

[0047] (3) The phase axis above the collapse column is missing or its continuity is poor. In areas with small-scale collapse columns, the phase axis above the main phase axis of the target coal seam often exhibits a significant reduction in local continuity. Specifically, this can manifest as a missing or discontinuous phase axis, along with a local attenuation of the corresponding amplitude energy. This phenomenon is due to the disturbance of the overlying strata caused by the collapse column, resulting in inconsistencies in time and phase between the originally stable reflected wave groups. Minor undulations or shifts occur at local reflection interfaces, causing the phase axis to exhibit fine-scale bending, local concavity, and irregular undulations on the cross-section.

[0048] Geometrically, this anomaly is most pronounced near the center of the collapse column, gradually transitioning outwards into a stable, continuously recovering reflection zone. Therefore, it serves as an important boundary clue for characterizing the influence range of the collapse column. When observed along the survey line, the changes in the phase axis—from continuous to discontinuous, then missing, and finally back to continuous—often correspond to the spatial distribution of the collapse column's influence from weak to strong and then back to weak. Unlike the "neat faulting" caused by typical faults, this anomaly emphasizes "discontinuity," meaning the phase axis may not form a clear displacement section, but rather exhibits intermittent interruptions and energy attenuation at multiple points and segments. Compared to the tuning effect caused by coal seam thinning, this anomaly is typically accompanied by a combination of characteristics, including decreased phase stability, local concavity, and enhanced disturbance, rather than simply amplitude changes.

[0049] Therefore, in the comprehensive anomaly response extraction process of this invention, "partial absence or deterioration of continuity of the same phase axis above the collapse column" is regarded as one of the key anomaly modes at the profile scale, which is used to provide reliable constraints when screening boundary closure and breakpoints, thereby improving the accuracy of collapse column boundary delineation.

[0050] (4) Delayed diffraction wave disappears As the diameter of the collapse column decreases, the propagation time of diffracted waves within the column shortens, leading to a reduction in the delay time. When the diameter of the collapse column becomes sufficiently small, the delay time shortens to the point where it becomes indistinguishable on the time profile, causing the diffracted wave and the delayed diffracted wave to overlap, or the delayed diffraction to disappear. In this case, the entire base of the collapse column can be approximated as a point of abrupt change. When seismic waves propagate to this point of abrupt change, only a single diffraction phenomenon will occur. Therefore, when the collapse column is very small, due to limitations in precision, the delayed diffracted wave becomes difficult to distinguish, and only the diffracted wave appears.

[0051] (5) The continuity of the in-phase axis decreases rather than completely becomes disordered. In areas with small-scale collapse columns, due to their scale being close to the lateral resolution limit of seismic waves, the seismic reflection wavefield typically does not exhibit the strong reflection disruption or complete disorder commonly seen in medium- and large-scale collapse columns. Instead, it displays a characteristic of overall stable phase continuity but significant local decreases. Specifically, the coal seam phase axis can still be traced macroscopically, but within the influence range of the collapse column, its phase stability deteriorates significantly, with slight shifts or short-distance interruptions of the phase axis appearing locally.

[0052] This type of phase continuity decrease typically manifests as the phase axis changing from a straight, regular shape to an uneven, loosely arranged state, and the phase correspondence between adjacent seismic traces weakens, but has not yet evolved into completely disordered reflections. This "weak disorder" characteristic reflects the disturbance effect of the collapse column interior and its edge fracture zone on the seismic wave propagation path and phase response, and is particularly closely related to the heterogeneity of the collapse filling material, local wave velocity variations, and multipath interference.

[0053] Unlike the complete destruction of the cofferdam axis caused by tectonic structures such as faults and highly fractured zones, the phase anomalies caused by small-scale collapse columns are characterized by the preservation of continuity, limited degree of disorder, and localized spatial distribution. They often occur simultaneously with local depressions, accompanying phase anomalies, or phase tailing. Therefore, a decrease in the continuity of the cofferdam axis, rather than complete disorder, can serve as one of the important criteria for identifying small-scale collapse columns.

[0054] (6) Biaxial features In areas with small-scale collapse columns, the phase axis of the target coal seam often exhibits a morphological transformation on seismic profiles, transitioning from a single axis (one phase axis) to a dual axis (two nearly parallel phase axes) and then back to a single axis. For example... Figure 2b As shown, the blue line represents the profile outline of the collapse column, the pink line represents the target layer, and the dashed line indicates the phase axis, which is a typical biaxial or triaxial characteristic position. Specifically, before entering the influence range of the collapse column, the main reflection phase axis is continuous and the waveform is stable, showing a typical single-axis response. When the profile crosses the boundary of the collapse column and approaches the central area, one or more additional phase axes that are approximately parallel to the main reflection appear near the main reflection event, forming a "biaxial" or even "triaxial" structure. This additional axis can be located above or below the main reflection and is often accompanied by phase broadening and local displacement. After leaving the influence range of the collapse column, the multiaxial structure gradually disappears and returns to a single-axis continuous form.

[0055] This "dual-axis" phenomenon essentially reflects the local medium inhomogeneity and wave field complexity caused by the collapse column: on the one hand, the collapse column and its fracture zone cause abrupt changes in reflected waves over short distances, forming multiple distinguishable phase bands; on the other hand, the impedance abrupt changes between the surrounding rock and the collapsed filling body, as well as the irregularity of the boundary geometry, easily induce local scattering and diffraction phenomena, resulting in the appearance of "additional axes" or "secondary axes" near the main phase axis. Unlike the overall drift of the phase axis caused by the slow thinning of coal seams or gentle tectonic undulations, the dual-axis (multi-axis) feature usually has strong spatial limitations, corresponds to the boundary ring zone, and is most significant when crossing the central area. Therefore, it can be regarded as one of the key anomaly patterns for identifying small-scale collapse column profiles.

[0056] (7) Accompanying phases appear above and below the fall column Within the influence zone of a small-scale collapse column, one or more "accompanying phases" can often be identified above and below (especially below) the phase axis of the target coal seam. For example... Figure 2c As shown, the blue line represents the profile outline of the collapse column, the pink line represents the target layer, and the dashed line indicates the phase axis, which is a typical location of the accompanying phase response. Accompanying phases typically do not form independent, stable, and traceable reflection interfaces, but rather exhibit phase anomalies superimposed near the main phase axis. Unlike regular wave groups formed by conventional inter-layer reflections, the appearance of accompanying phases is localized and unstable. Their spatial location often corresponds to the geometric disturbance center or boundary transition zone of the collapse column, and they tend to repeat on seismic profiles in different azimuths.

[0057] From the perspective of wavefield mechanisms, the localized media fracturing, microcrack development, and rapid lateral changes in wave impedance caused by small-scale collapsed columns enhance scattering and diffraction components, resulting in additional energy changes near the main reflected wave. Simultaneously, these accompanying phases reflect more the reduced stability of wavefield propagation and waveform distortion caused by structural disturbances than simple amplitude anomalies. Therefore, even if the amplitude contrast within the anomalous region is not significant, the accompanying phases remain clearly visible, greatly enhancing their applicability.

[0058] In terms of morphological appearance, the accompanying phase anomaly bands are usually distributed along the main phase axis, exhibiting a "band-like, discontinuous" characteristic: the phase anomaly is most concentrated near the center of the affected area of ​​the collapsed column, gradually weakening outwards and transitioning to normal phase; one or more anomaly bands may appear above and below the target phase axis, and in some cases, they may even locally overlap with the main phase axis, causing the phase axis to appear "widened" or "ghosted." It is important to emphasize that this anomaly band is often more prominent below the collapsed column, possibly due to disturbances in the medium below the collapsed column and the accumulation of scattered energy, which makes the phase stability of the underlying wave field decrease more significantly.

[0059] Near the fault crater, the primary reflection event often exhibits an asymmetric response pattern: the phase axis on the side closer to the fault crater may show a more pronounced break, faulting, or phase abrupt change, while the other side remains weakly continuous or only shows enhanced bending. At the same time, phase tailing is more easily enhanced on the break or phase abrupt change side, manifested as a longer extension of phase energy along the time direction and a more "elongated" waveform. This asymmetry of "break on one side and weak continuity on the other" differs from the continuous bending characteristics of direct faulting or folding in general faults.

[0060] In this embodiment of the invention, the combined feature of "accompanying phase + asymmetric response at fault point + phase tail" is incorporated into the comprehensive anomaly response mode for: firstly, to determine the development location of potential collapse columns at the profile scale; and secondly, to jointly verify with geometric perturbation zones and wavefield uniformity reduction zones at the planar scale, thereby providing a stable and repeatable basis for determining the central control point and screening the boundary control point. This feature is not sensitive to amplitude contrast and can maintain good recognizability even when amplitude changes are not significant, thus it can be used as one of the important criteria for distinguishing small-scale collapse columns from general fault and fold disturbances.

[0061] II. Planar features include: (1) Local dip morphological disturbance: By analyzing the local variation characteristics of the geometric morphology in the slice along the target layer, small-scale collapse columns are abnormal regions with sudden dip angle changes, nonlinear disturbances, directional discrepancies or value range changes in a small area. The boundaries of abnormal dip angle and azimuth angle changes are usually used as the boundaries of collapse columns. This type of local dip morphological disturbance is used as an important planar marker for the identification of small-scale collapse columns.

[0062] (2) Reduced wavefield consistency: By analyzing the consistency changes of 3D seismic data on layer slices, we can identify anomalous zones that are characterized by reduced waveform similarity, weakened continuity, weaker reflected wave amplitude than normal standard reflected energy, and concentrated spatial distribution. These anomalous zones have obvious responses in amplitude and variance properties on layer slices and can be interpreted as suspected small-scale collapse column boundaries.

[0063] (3) Transformation of the orderly to disordered reflective texture: By analyzing the directionality, continuity and overall morphological characteristics of the seismic reflective texture features on the plane, local abnormal areas that transform from ordered to disordered textures are identified, and such texture destruction phenomena are used as an important basis for identifying the influence area of ​​small-scale collapse columns.

[0064] By comprehensively analyzing the spatial superposition relationship of various planar anomaly response characteristics, when multiple anomalies correspond to each other in spatial location or show a consistent distribution trend, the anomaly region is identified as a small-scale collapse column.

[0065] In step S3, the abnormal concentration area and boundary ring zone are identified; Based on the spatial distribution and continuity characteristics of the comprehensive anomaly response on a planar scale, the boundary zone between the stable region and the disturbance region of the anomaly response is identified.

[0066] In practical implementation, the location where abrupt changes in anomaly intensity, geometry, or stability indicators occur can be identified by analyzing the changing trends of the comprehensive anomaly response on the target stratum plane. This location is then considered the spatial boundary zone where the anomaly transitions from the stable zone to the disturbed zone. Furthermore, this spatial boundary zone is constructed as a closed or semi-closed boundary ring to initially constrain the influence range of potential collapse columns on the target coal seam.

[0067] In one implementation, the target stratigraphic plane is considered as a spatial field of anomalous responses. Within this spatial field, the aggregation patterns and distribution continuity characteristics of multi-source anomalous responses, which characterize geometric changes, wavefield properties, and the stability of reflection structures, are comprehensively analyzed on the plane. By identifying the spatial boundary zones formed when anomalous responses transition from stable to disturbed regions, the anomalies are transformed from "discrete features" into "boundary structures with confinement significance."

[0068] When the aforementioned anomalous response exhibits point-like, ring-like, or semi-ring-like characteristics on a planar scale and possesses clear internal and external spatial differences, this continuous distribution area is defined as the boundary ring of a potential collapse column. The boundary ring is determined not by whether the anomalous response is completely closed, but by its effective spatial constraint on the anomalous body.

[0069] This approach enables a shift from traditional "identification of anomalies or anomalies" to "extraction of anomaly spatial organization structure," allowing the planar influence range of a collapsed column to be stably delineated without relying on a clear structural interface, and providing a unified spatial constraint framework for subsequent determination of the center location and multi-directional profile interpretation.

[0070] The specific process of step S3 is as follows: Step S31: Identification of Planar Anomaly Response Pattern of Collapse Column The comprehensive anomaly response obtained in step S2 is analyzed on the target stratigraphic plane or along the stratigraphic slice to identify the typical response patterns of small-scale collapse columns on the planar scale.

[0071] The planar anomaly response forms include, but are not limited to, one or more of the following: (1) Local geometric disturbance forms: manifested as abrupt changes or gradient enhancement of dip angle, azimuth angle or stratigraphic morphology within a small range; (2) Forms of reduced wave field uniformity: manifested as decreased waveform similarity, weakened amplitude stability, or reflection texture changing from regular to discrete; (3) Changes in the stability of the reflective structure: The reflective texture changes from ordered to disordered, but no obvious break or interruption of reflection is formed.

[0072] The aforementioned abnormal responses typically appear as point-like or patchy distributions, and spatially they tend to concentrate around the development area of ​​the collapse column.

[0073] Step S32: Spatial distinction between stable and disturbed regions Based on the identification of the anomaly response patterns in the plane, the target stratigraphic plane is divided into the following two types of regions: Stable region: A region where the abnormal response value changes gently, the wave field structure is continuous, and the reflection pattern is regular; Disturbance zone: The region where abnormal response values ​​change significantly, wavefield uniformity decreases, and reflection structure shows a trend of disorder.

[0074] By analyzing the magnitude of numerical changes, gradient characteristics, or spatial continuity of abnormal responses on a plane, the spatial distribution range of stable and disturbed regions can be determined.

[0075] Step S33: Identification of transition features from stable region to disturbed region Between the stable and perturbation regions, further identify transitional features of the anomalous response from a stable state to a perturbation state. These transitional features include, but are not limited to: Abnormal response values ​​undergo abrupt changes within a small spatial scale; The gradient transitions rapidly from the low-value region to the high-value region; Spatial continuity changes from a regular form to an irregular form.

[0076] This transitional zone typically corresponds to the spatial boundary where the influence of a collapse column increases from weak to strong, and is an important planar marker for depicting the extent of the collapse column's influence.

[0077] Step S34: Construction of Spatial Boundary Zone Based on the transition features identified in step S33, continuous or near-continuous anomalous change trajectories are extracted along the boundary between the stable and disturbed regions within the target stratigraphic plane, and these trajectories are constructed as spatial boundary zones.

[0078] The spatial boundary zone can be closed or semi-closed, used to characterize the effective influence range of potential collapse columns, and as a geometric constraint for subsequent collapse column center location and boundary interpretation.

[0079] The specific process of S34 is as follows: Step S341: Identification of Planar Anomaly Gradients and Transition Regions Within the target stratigraphic plane, the comprehensive anomaly response is digitally represented, and its variation characteristics at the planar scale are calculated, preferably including one or more of the following methods: Anomaly intensity gradient analysis: Calculate the spatial gradient or rate of change of the comprehensive anomaly response on the plane to characterize the transition position of the anomaly response from the stable region to the disturbance region; Analysis of abnormal continuity changes: Analyze the continuity, integrity or order of abnormal responses on a plane to identify areas where continuity is significantly reduced or disrupted; Anomaly Consistency Mutation Analysis: Planar analysis is performed on wavefield consistency or structural stability-related properties to identify locations where abrupt changes or rapid decays occur from high consistency to low consistency.

[0080] Regions that meet one or a combination of the above conditions are identified as abnormal transition zones.

[0081] Step S342: Transition Zone Trajectory Extraction Within the identified abnormal transition region, continuous or near-continuous abnormal change trajectories are extracted along the spatial contact position between the stable region and the disturbance region. The trajectory extraction preferably includes: Extreme value or abrupt change line extraction: Extract abnormal change trajectories along the abnormal gradient where local extreme values ​​are reached or the rate of change increases significantly; Contour extraction: Based on the abnormal response intensity or consistency index, a threshold is set, and the corresponding contour lines are extracted as the transition zone trajectory; Continuity disruption boundary extraction: Extract the geometric boundary of the region where the abnormal continuity is significantly reduced or interrupted, forming an abnormal change trajectory.

[0082] The above methods can be used individually or in combination to improve the stability of transition zone trajectory extraction.

[0083] Step S343: Construction and Morphological Constraints of Spatial Boundary Zones The abnormal change trajectories extracted in step S342 are spatially connected, smoothed, and morphologically constrained on a plane to construct spatial boundary zones, specifically including: Trajectory connection and completion: For abnormal trajectories with local missing parts or discontinuities, connect them according to their spatial direction and trend of change; Geometric smoothing and simplification: The trajectory is smoothed to eliminate high-frequency oscillations caused by noise or local anomalies, making its shape more consistent with geological understanding; Closure determination: Determine whether the trajectory forms a closed loop; if it does not form a complete closure, retain it as a semi-closed shape and record its opening direction and position.

[0084] Step S344: Boundary loop generation and constraint assignment Centered on the spatial boundary zone obtained in step S343, a boundary ring with a certain width is constructed to characterize the effective influence range of the potential collapse column on a planar scale. The process includes: Ring width determination: The width range of the boundary ring is determined based on the gradient change in the abnormal response transition zone; Annular region generation: Closed or semi-closed boundary annular regions are formed by the expansion on both sides of the spatial boundary zone; Geometric constraint assignment: The boundary ring is used as the geometric constraint condition for subsequent center positioning of the collapse column (step S4) and boundary interpretation (steps S5–S7).

[0085] In step S4, the central control point is determined; Within the constraints of the aforementioned closed or semi-closed boundary ring, multiple seismic profiles passing through the boundary ring are selected, and the positions corresponding to the integrated anomaly response space in each profile are analyzed to extract candidate center positions.

[0086] Subsequently, spatial consistency analysis was performed on the candidate center locations obtained from seismic profiles in different orientations. For example, the degree of spatial concentration or frequency of recurrence of candidate locations was compared, and the location with the highest consistency was selected as the center control point of the collapse column, thereby improving the reliability of center location determination.

[0087] In one embodiment, within the constraint range of the closed or semi-closed boundary ring, multiple seismic profiles passing through the boundary ring are selected, and candidate center positions corresponding to the integrated anomaly response space in each profile are extracted. The spatial consistency of the candidate center positions in different orientation profiles is analyzed, and the position with the highest consistency is determined as the center control point of the collapse column.

[0088] In each of the aforementioned seismic profiles, the possible center location of the collapse column is determined based on the comprehensive profile anomaly response extracted in step S2. The comprehensive profile anomaly response includes, but is not limited to: Local disturbances in the phase axis, exhibiting a "concave" phenomenon; phase tailing; partial absence or deterioration of the phase axis above the collapse column; disappearance of delayed diffraction waves; decreased rather than complete disruption of the phase axis continuity; biaxial characteristics; accompanying phase below the collapse column. By comparing the spatial locations of the above anomalous responses on the cross-section, candidate locations of potential collapse column centers on each cross-section are determined.

[0089] Spatial consistency analysis was performed on the candidate center locations from different azimuth profiles. Regions that repeatedly appear across multiple profiles and exhibit the smallest spatial deviation were preferentially selected as the center control points for the collapsed columns. When there is a certain offset between the center locations determined from different profiles, the point with the most concentrated and stable anomaly response characteristics was chosen as the center location, thereby reducing the impact of a single profile or local anomaly on the center location results.

[0090] The specific process of step S4 is as follows: Step S41: Selection of seismic profiles traversing the boundary ring In the target horizon plan view window, using the closed or semi-closed boundary ring constructed in step S3 as a constraint, multiple seismic profiles passing through the boundary ring are automatically selected. The selection method preferably includes, but is not limited to: (1) Azimuth coverage principle: The profile azimuth should cover different azimuth directions, preferably with uniform azimuth distribution, in order to reduce the error caused by a single survey line azimuth; (2) Crossing constraint principle: Each profile should intersect the boundary ring in the plane projection and have a sufficiently long crossing section within the ring to ensure that the profile can sample the disturbance area within the ring; (3) Quality optimization principle: When there are multiple candidate profiles, the profile with a high signal-to-noise ratio of target reflection event and good trackability of the same phase axis should be selected. (4) Quantity constraint: It is preferred to select no less than three profiles, and even more preferably 6-12 profiles, so as to balance azimuth coverage and work efficiency.

[0091] In one implementation, a radial profile can be generated at preset azimuth intervals, with the geometric center of the boundary ring as a reference, to achieve multi-directional cross-sampling of the boundary ring.

[0092] Step S42: Extraction of candidate center locations For each selected seismic profile, within the neighborhood above and below the target horizon, and in conjunction with the comprehensive anomaly response model from step S2, candidate center locations corresponding to the potential collapse column development locations are extracted. The extraction of candidate center locations preferably includes, but is not limited to, the following methods: (1) Abnormal window limitation: The projection range of the boundary ring on the plane is mapped onto the profile, limiting the candidate center to be searched only within the ring constraint range; (2) Anomaly indication feature extraction: Automatically identify the profile anomaly location corresponding to the comprehensive anomaly response within the search range. The anomaly location can be determined by the characteristic manifestations of reflection structure disturbance, wave field uniformity reduction, or phase and structure joint stability change. (3) Determination of candidate center points: The above-mentioned abnormal locations are expressed in a centralized manner to obtain candidate center locations; among them, the candidate center locations can be the point with the strongest abnormal response, the geometric center of the abnormal response area, or the symmetric center of the abnormal response area along the cross-sectional direction.

[0093] In a preferred implementation, the candidate center position is determined as: a spatial position where the abnormal response intensity reaches an extreme value near the target reflection event and has a continuous corresponding position in the upper and lower neighborhoods.

[0094] Step S43: Spatial consistency analysis of multi-directional candidate center locations Candidate center locations extracted from different profiles are uniformly mapped onto the target stratigraphic plane, and the spatial consistency of the candidate center locations is analyzed to determine the final center control point. The consistency analysis preferably includes, but is not limited to: (1) Clustering analysis: Calculate the spatial clustering degree of the candidate center locations on the plane, and select the clustering center with the highest clustering degree as the central control point; (2) Frequency of recurrence analysis: The number of times the candidate center location falls into the same spatial neighborhood in different orientation profiles is counted, and the spatial neighborhood center with the most occurrences is selected as the center control point; (3) Dispersion constraint: Calculate the distance distribution from the candidate center location to the cluster center. If the dispersion exceeds the preset threshold, return to step S42 to review the abnormal indication features or adjust the profile selection. (4) Confidence-based selection rule: When multiple candidate cluster centers exist at the same time, the cluster center with higher comprehensive abnormal response intensity in step S2 and more consistent spatial correspondence within the boundary ring in step S3 is selected as the central control point.

[0095] Finally, the candidate center location with the highest consistency is determined as the center control point of the collapse column, which is used for multi-directional boundary sampling and boundary control point extraction in the subsequent step S5.

[0096] In step S5, candidate boundary control points are obtained; Using the central control point as a reference, multiple seismic profiles are automatically constructed within the target horizon plane at preset azimuth intervals. In this embodiment, the azimuth interval can be an equal-angle interval, preferably 30°.

[0097] In each of the aforementioned seismic profiles, anomaly locations corresponding to the integrated anomaly response near the central control point are identified, and these locations are used as candidate boundary control points, thereby enabling multi-directional sampling of potential collapse column boundaries.

[0098] In one implementation, using the central control point of the collapse column determined in step S4 as a reference, a set of multi-directional radial profiles is automatically generated around the central control point in the planar base map window of the target coal seam. Specifically, using the central control point as the geometric starting point, seismic profiles passing through the central control point are constructed in the plane at preset azimuth angle steps (preferably every 30°, but other equal or non-equal intervals are also possible). This allows the same potential anomaly to be repeatedly sampled in different propagation directions, reducing the risk of boundary drift and missed detection caused by the azimuth offset of a single survey line.

[0099] In each of the aforementioned multi-directional seismic profiles, the neighborhood of the central control point is used as the starting search range to automatically identify and locate the comprehensive anomaly response associated with small-scale collapse columns. The comprehensive anomaly response includes at least one or more of the following: local disturbance of the in-phase axis, appearance of a "concave" phenomenon; phase tailing; partial absence or deterioration of the in-phase axis above the collapse column; disappearance of delayed diffraction waves; decreased continuity of the in-phase axis rather than complete disorder; accompanying phase; and biaxial characteristics.

[0100] The system scans abnormal change points along the profile from the center to both sides, extracts abnormal locations that meet the discrimination rules, and uses them as candidate boundary control points. Among them, the candidate boundary control points are preferably located at the locations with the most obvious features within a certain distance of the central control point, so as to ensure that their corresponding collapse column boundaries are met.

[0101] Finally, the candidate boundary control points obtained from multiple profiles in different orientations form a set of points with multi-directional constraints, providing boundary control conditions for subsequent planar projection, linkage verification, and closure.

[0102] The specific process of step S5 is as follows: Step S51: Automatic construction of multi-azimuth seismic profiles Using the center control point of the collapse column determined in step S4 as the geometric reference, multiple seismic profiles passing through the center control point are automatically constructed in the target horizon plan view window. The construction method of the seismic profiles preferably includes, but is not limited to, the following steps: (1) Establish a polar coordinate reference system: take the position of the central control point in the target layer plane as the origin and establish a planar polar coordinate system; (2) Azimuth angle preset: Several azimuth angle directions are preset in the polar coordinate system, and radial seismic profiles passing through the central control point are generated along each azimuth angle direction; (3) Profile length setting: The length of each seismic profile on the plane should at least cover the closed or semi-closed boundary ring constructed in step S3, and extend appropriately inward and outward to ensure that the response characteristics of the anomaly transitioning from the stable zone to the disturbed zone can be fully sampled; (4) Profile generation method: The seismic profile can be automatically generated by a pre-set azimuth interval.

[0103] Step S52: Principles for setting azimuth intervals The preset azimuth interval is determined comprehensively based on the scale characteristics of potential collapse columns, the lateral resolution of seismic data, and computational efficiency, preferably following the principles below: (1) Direction coverage principle: The azimuth angle should cover the entire 360° spatial direction to avoid boundary identification deviation due to a single or few azimuths; (2) Resolution matching principle: The azimuth interval should be matched with the lateral resolution of the seismic data and the typical scale of the collapse column, so that the sampling interval between adjacent profiles on the plane is smaller than the minimum identifiable scale of the potential collapse column. (3) Efficiency and accuracy balance principle: Under the premise of ensuring the stability of boundary recognition, avoid redundant calculations caused by excessively dense azimuth angles.

[0104] In a preferred embodiment, the azimuth interval is an equal angular interval, more preferably 30°, that is, 12 seismic profiles are constructed within the range of 0°-360°; when higher accuracy is required, the azimuth interval can be reduced to 15°, and it can also be widened to 45° in the rapid screening stage.

[0105] Step S53: Limiting the search for anomalous response locations in the profile For each constructed seismic profile, within the target horizon and its adjacent time windows, using the central control point as a reference, anomaly response location searches are performed in its surrounding area. The search preferably includes: (1) Search range limitation: Along the seismic profile direction, starting from the central control point, search to the profile projection position corresponding to the closed or semi-closed boundary ring on both sides; (2) Abnormal response criterion invocation: Within the search range, in combination with the comprehensive abnormal response mode established in step S2, identify the profile locations where there are significant local disturbances, wave field consistency changes, or phase and structural joint stability changes in the reflection structure. (3) Boundary candidate location determination: The location where the abnormal response changes from a stable state to a significant disturbance state, or the location where the abnormal response intensity changes abruptly or reaches an extreme value, is taken as the potential boundary response location.

[0106] Step S54: Extraction of candidate boundary control points Based on the abnormal response locations identified in step S53, one or more candidate boundary control points are extracted on each seismic profile. The extraction method preferably includes, but is not limited to: (1) Anomaly initiation point method: The location where the anomaly response first appears significantly is taken as the candidate boundary control point; (2) Abnormal extreme value method: The location where the abnormal response intensity reaches the local extreme value is taken as the candidate boundary control point; (3) Abnormal transition zone center method: When the abnormal response presents a continuous transition zone, the geometric center of the transition zone in the profile direction is taken as the candidate boundary control point.

[0107] Each seismic profile can obtain two candidate boundary control points. All candidate boundary control points retain their corresponding spatial coordinate information for the plane-profile joint constraint verification in the subsequent step S6.

[0108] In step S6, joint constraint verification and effective boundary point set formation are performed; The aforementioned candidate boundary control points are uniformly mapped to the target layer plane, and the candidate boundary control points are screened and corrected by combining the spatial consistency characteristics of the comprehensive anomaly response.

[0109] Specifically, it can be determined whether the distribution of candidate boundary control points on the plane corresponds to the spatial morphology of closed or semi-closed boundary rings; for control points that do not meet the spatial consistency constraints, they can be automatically corrected or manually returned to the corresponding seismic profile to re-identify the anomalous response locations. After joint constraint verification, a set of effective boundary points that conforms to geological understanding is formed.

[0110] In one implementation, the candidate boundary control points are uniformly mapped to the target layer plane, and combined with the spatial consistency characteristics of the comprehensive anomaly response, candidate boundary control points that do not meet the preset spatial constraints are eliminated or corrected to form an effective boundary point set.

[0111] Within the aforementioned stratigraphic plane base map window, the spatial rationality of candidate boundary control points is jointly verified by combining the analysis results of multiple seismic attributes. The seismic plane characteristic analysis results are used to characterize local geometric changes, changes in wavefield propagation consistency, and changes in the stability of reflection structures. By analyzing the spatial correspondence between candidate boundary control points and the aforementioned planar anomaly areas, it is determined whether the candidate boundary control points are located within a reasonable range of the potential collapse column boundary.

[0112] For candidate boundary control points that significantly deviate from the local geometric disturbance zone, wavefield consistency reduction zone, or reflection texture anomaly zone on the plane, they are determined not to meet the spatial consistency constraint conditions and are automatically eliminated or corrected according to preset rules. The correction includes automatic elimination correction or manual return to the corresponding seismic profile to re-examine the anomalous response location, or adjustment of the control point location within the allowed neighborhood range.

[0113] After the above screening and correction, the remaining candidate boundary control points exhibit good continuity and reasonable spatial distribution on the plane, forming a set of effective boundary points that conforms to the actual geological characteristics. This set of effective boundary points provides reliable geometric constraints for subsequent boundary closure and determination of the planar morphology of collapse columns.

[0114] In step S7, the boundary contour is generated; Based on the effective boundary point set, the planar boundary profile of the collapse column is generated in the target stratigraphic plane through an adaptive boundary fitting method.

[0115] In practice, the boundary profile can be locally smoothed based on the distribution density changes of effective boundary points on the plane, thereby obtaining a continuous and stable closed or nearly closed boundary, which can be used to characterize the effective influence range of the collapse column on the target coal seam.

[0116] In one embodiment, based on the effective boundary point set, the planar boundary profile of the collapse column is generated in the target stratum plane by an adaptive boundary fitting method, thereby obtaining the effective influence range of the collapse column on the target coal seam.

[0117] During the boundary fitting process, the boundary fitting parameters are adaptively adjusted based on the distribution density and spatial location characteristics of the effective boundary points on the plane. This ensures that the fitting result can both reflect the overall spatial control trend of the effective boundary point set and maintain responsiveness to changes in boundary morphology in local regions. The adaptive boundary fitting preferably employs one or more of curve smoothing and local densification methods, but is not limited to specific mathematical models or implementation algorithms.

[0118] Through the above fitting process, continuous, closed or nearly closed collapse column boundary profiles are generated to characterize the effective influence range of the collapse column at the target layer.

[0119] The boundary of the collapse column serves as the basic input for subsequent three-dimensional spatial morphological constraints and visualization verification, providing reliable geometric boundary conditions for the fine depiction of the spatial relationship between the collapse column and the target coal seam.

[0120] The specific process of step S7 is as follows: The specific process of S7 is as follows: Step S71: Spatial sorting of boundary points and establishment of connectivity The effective boundary point set is spatially organized within the target stratigraphic plane, specifically including: Polar angle or azimuth sorting: Using the central control point determined in step S4 as a reference, sort each effective boundary point according to the azimuth order relative to the central control point. Neighborhood connectivity construction: Based on the spatial distance, directional continuity and anomaly response consistency between adjacent boundary points, the initial connectivity between boundary points is established to form the initial boundary skeleton.

[0121] The above processing ensures that the connection order of boundary points on the plane is consistent with the actual geological spatial relationship.

[0122] Step S72: Boundary point density and stability assessment The distribution characteristics of the effective boundary point set on the plane are evaluated to determine the adaptive adjustment criteria for boundary fitting, specifically including: Point density analysis: Calculate the distribution density of boundary points in different orientations or spatial segments, and identify dense and sparse areas of boundary points; Stability evaluation: The reliability of boundary points is assessed by combining the spatial consistency characteristics of the comprehensive anomaly response, and stable boundary points are distinguished from unstable boundary points.

[0123] The above evaluation results serve as the basis for local densification or smoothing in the subsequent boundary fitting process.

[0124] Step S73: Adaptive boundary curve fitting Based on steps S71 and S72, adaptive curve fitting is performed on the boundary point set, specifically including: Local densification fitting: In sections with high boundary point density and strong consistency of anomalous responses, higher fitting accuracy is used to maintain the sensitivity of the boundary contour to local geometric changes. Local smoothing fitting: In sections where boundary points are sparse or stability is low, the fitting complexity is reduced and the boundary contour is smoothed to avoid unreasonable boundary fluctuations caused by noise or isolated anomalies. Continuity and closure constraints: During the fitting process, continuity constraints are applied to the boundary curves, and closed planar boundary contours are generated based on the distribution of boundary points.

[0125] The adaptive process achieves automatic adjustment of boundary contours to data quality and anomaly stability by employing different fitting strategies for different segments.

[0126] Step S74: Boundary contour rationality check The generated planar boundary profile is checked for reasonableness, specifically including: Geometric consistency check: Determine whether the boundary profile is consistent with the spatial morphology of the closed or semi-closed boundary ring constructed in step S3; Verification of the correspondence between abnormal responses: Verify whether the boundary contour falls entirely within the spatial range of the transition from the stable region to the disturbed region in the comprehensive abnormal response; Iterative correction: Adjust local boundary segments that do not meet the above conditions until the constraints of planar anomaly distribution and geological cognition are met.

[0127] In step S8, 3D visualization verification is performed; Finally, the spatial relationship between the generated planar boundary contour and the target coal seam is verified using three-dimensional visualization.

[0128] Specifically, the plane boundary contour of the collapse column, the target coal seam level, and the corresponding seismic profile can be jointly displayed in three-dimensional space to verify the spatial correspondence between the boundary contour and the comprehensive anomaly response, thereby further confirming the rationality of the prediction results.

[0129] In one implementation, after obtaining the closed boundary of the small-scale collapse column, the boundary result, along with the target coal seam horizon, relevant seismic data volume, and seismic attribute volume, is loaded into a three-dimensional visualization environment to automatically construct a three-dimensional spatial relationship model between the collapse column and the coal seam. By comprehensively observing the relative positional relationship, interpenetration relationship, and continuity changes of the collapse column boundary and the coal seam reflection morphology in three-dimensional space, the consistency of the boundary with the seismic profile and planar anomaly response is verified.

[0130] During visualization, the collapse column boundary is rotated, cut, and scaled at different orientations and scales to examine the disturbance range, boundary integrity, and spatial closure rationality of the collapse column to the target coal seam. When a mismatch is found between the local boundary and the coal seam's reflection morphology or attribute anomaly, the previous steps are returned to correct the relevant boundary control points. Through this three-dimensional visualization verification method, the interpretation results of the collapse column boundary are comprehensively verified, further reducing the uncertainty caused by planar or single-section interpretation and improving the reliability and engineering application value of small-scale collapse column prediction results.

[0131] Based on the same inventive concept, embodiments of the present invention also provide a rapid prediction system for small-scale collapse columns in coal mines, such as... Figure 3 As shown, it includes: The stratigraphic construction module generates the target coal seam stratigraphic level based on the well-seismic calibration results and combined with the automatic recursive method of spatial small-area elements. The anomaly response analysis module analyzes the three-dimensional seismic data of the target layer and its neighborhood to extract the comprehensive anomaly response used to characterize potential small-scale collapse columns. An anomaly concentration area identification module identifies the spatial boundary zone of the comprehensive anomaly response on a planar scale and constructs a closed or semi-closed boundary ring for constraining the influence range of potential collapse columns. The central control point determination module performs spatial consistency analysis on candidate center locations in multiple seismic profiles within the boundary ring constraint range to determine the central control point of the collapse column. The multi-directional interpretation module constructs multi-directional seismic profiles within the target horizon plane and obtains candidate boundary control points based on the central control point. The joint constraint verification module maps candidate boundary control points to the target layer plane and performs screening and correction based on the spatial consistency characteristics of the comprehensive anomaly response. The boundary generation module generates the planar boundary profile of the collapsed column based on the effective boundary point set through an adaptive boundary fitting method. The 3D verification module verifies the spatial relationship between the planar boundary contour and the target coal seam through 3D visualization.

[0132] In this embodiment of the invention, a 1.33km section of the Pingxing Company mining area in Yangquan, Shanxi Province was selected. 2 And one mine, 6.8km 2 Using 3D seismic data as a field application example, the proposed rapid prediction method for small-scale collapse columns was validated. The comprehensive anomaly response results obtained based on the aforementioned actual data are as follows: Figure 2d and Figure 2eAs shown in the figure. The results show that the present invention can effectively identify the comprehensive anomaly response characteristics of small-scale collapse columns in in-situ 3D seismic data, and achieve a fine characterization of their influence range and boundary morphology, thereby verifying the practicality and reliability of the method of the present invention.

[0133] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to the method section.

[0134] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A rapid prediction method for small-scale collapse columns in coal mines, characterized in that, Includes the following steps: S1. Based on the well-seismic calibration results, determine the seismic phase axis corresponding to the target coal seam, and use the automatic recursive method of spatial small-area elements to interpret the stratigraphic position and generate the target stratigraphic position. S2. Analyze the three-dimensional seismic data within the target layer and its neighborhood to establish a comprehensive anomaly response. The comprehensive anomaly response characterizes the local disturbance of the reflecting structure, wavefield consistency and phase and structural joint stability changes, and delineates the seismic response characteristics of potential small-scale collapse columns. S3. Based on the spatial distribution and continuity characteristics of the comprehensive anomaly response on a planar scale, identify the spatial boundary zone transitioning from the stable region to the disturbance region, and construct the spatial boundary zone into a closed or semi-closed boundary ring. S4. Within the constraint range of the closed or semi-closed boundary ring, select multiple seismic profiles, extract the candidate center positions in each seismic profile that correspond to the spatial relationship of the integrated anomaly response, and analyze the spatial consistency of the candidate center positions in different orientation profiles to determine the position with the highest consistency as the center control point of the collapse column. S5. Using the central control point as a reference, construct multiple seismic profiles at preset azimuth intervals within the target layer plane, identify anomalous response locations near the central control point in each seismic profile, and obtain multiple candidate boundary control points. S6. Map the candidate boundary control points to the target layer plane in a unified manner, and combine the spatial consistency characteristics of the comprehensive anomaly response to eliminate or correct candidate boundary control points that do not meet the preset spatial constraints, so as to form an effective boundary point set. S7. Based on the effective boundary point set, generate the planar boundary profile of the collapse column in the target stratum plane through an adaptive boundary fitting method, and obtain the influence range of the collapse column on the target coal seam. S8. Verify the spatial relationship between the planar boundary contour and the target coal seam using three-dimensional visualization.

2. The rapid prediction method for small-scale collapse columns in coal mines as described in claim 1, characterized in that, The method for obtaining the target layer in step S1 is as follows: Using a manually calibrated sample profile as an initial constraint, the sample profile is divided into several three-dimensional micro-elements. The target layer is recursively generated in three-dimensional space based on the waveform feature comparison method, while allowing manual correction of the recursion process in real time.

3. The rapid prediction method for small-scale collapse columns in coal mines as described in claim 1, characterized in that, The profile anomaly modes of the comprehensive anomaly response described in step S2 include: local disturbances and multi-layer "concave" deformation features of the in-phase axis; multiple accompanying phase features above and below the in-phase axis in the collapse column development area; features of local absence or reduced continuity of the in-phase axis; features of weakened or disappeared delayed diffraction wave response; features of decreased phase continuity of reflected waves but without complete disorder; features of the in-phase axis transitioning from a single-axis response to a dual-axis or multi-axis response and rapidly recovering; and features of phase tailing at the boundary of collapse column development.

4. The rapid prediction method for small-scale collapse columns in coal mines as described in claim 1, characterized in that, The comprehensive abnormal response mentioned in step S2 is a combination of one or more of the following characteristics: accompanying phase response appears above and below the phase axis, while the phase continuity decreases but does not become completely disordered, and the phase axis changes from a single-axis response to a biaxial or multiaxial response. This is used as a criterion for distinguishing small-scale collapse columns from faults, folds or coal seam scour zones.

5. The rapid prediction method for small-scale collapse columns in coal mines as described in claim 1, characterized in that, In step S3, the closed or semi-closed boundary ring is formed by identifying the gradient change characteristics of the integrated anomaly response on a planar scale, wherein the gradient change characteristics include the abrupt change characteristics of the anomaly from the stable region to the perturbation region.

6. The rapid prediction method for small-scale collapse columns in coal mines as described in claim 1, characterized in that, In step S3, the closed or semi-closed boundary ring is determined by the spatial superposition relationship of multiple types of integrated anomalous responses, which include at least reflection structure disturbance anomalies and wave field consistency change anomalies.

7. The rapid prediction method for small-scale collapse columns in coal mines as described in claim 1, characterized in that, In step S4, the spatial consistency analysis of the candidate center location includes comparing the spatial concentration or recurrence frequency of abnormal locations in different orientation profiles.

8. The rapid prediction method for small-scale collapse columns in coal mines as described in claim 1, characterized in that, In step S6, the constraint of spatial consistency includes: the distribution of candidate boundary control points on the target stratigraphic plane must correspond to the spatial morphology of the closed or semi-closed boundary ring.

9. A rapid prediction method for small-scale collapse columns in coal mines as described in claim 1, characterized in that, In step S7, the adaptive boundary fitting method smooths the boundary contour based on the distribution density change of the effective boundary point set on the plane.

10. A rapid prediction system for small-scale collapse columns in coal mines, characterized in that, include: A stratigraphic construction module, which generates the target coal seam stratigraphic level based on well seismic calibration results and combined with an automatic recursive method using small spatial elements; An anomaly response analysis module analyzes three-dimensional seismic data within the target layer and its neighborhood to extract a comprehensive anomaly response for characterizing potential small-scale collapse columns. An anomaly concentration area identification module identifies the spatial boundary zone of the integrated anomaly response on a planar scale and constructs a closed or semi-closed boundary ring for constraining the influence range of potential collapse columns. The central control point determination module performs spatial consistency analysis on candidate center locations in multiple seismic profiles within the boundary ring constraint range to determine the central control point of the collapse column. A multi-directional interpretation module, which constructs a multi-directional seismic profile within the target horizon plane and obtains candidate boundary control points based on the central control point; The joint constraint verification module maps candidate boundary control points to the target layer plane and performs screening and correction based on the spatial consistency characteristics of the comprehensive anomaly response. A boundary generation module generates the planar boundary profile of the collapsed column based on an effective set of boundary points using an adaptive boundary fitting method. The three-dimensional verification module verifies the spatial relationship between the planar boundary contour and the target coal seam through three-dimensional visualization.

Images