A karst goaf identification method based on multi-source data fusion

Abstract

The application discloses a karst goaf recognition method based on multi-source data fusion, relates to the technical field of geotechnical engineering investigation and bridge foundation engineering, and takes a bridge pile foundation influence body as a karst goaf recognition object, so that the bridge pier platform, pile body, pile end bearing layer and adjacent pile correlation range are included in the same three-dimensional engineering recognition unit, a corresponding relationship is formed among the abnormal positions, abnormal layer positions and engineering objects of multi-source geological data, cause evidence records are generated in each bridge pile foundation influence body, and karst cause evidence, goaf cause evidence, hydrological disturbance evidence and shallow disturbance evidence are recorded respectively, so that different abnormal sources can be distinguished within the same pile foundation influence range, geophysical prospecting anomalies, drilling exposure, mining data, remote sensing deformation, hydrological information and construction process information are corresponded to specific cause categories, and the disposal basis switching caused by the confusion of karst causes, goaf causes and hydrological disturbance causes is reduced.

Description

Technical Field

[0001] This invention relates to the fields of geotechnical engineering investigation and bridge foundation engineering technology, and in particular to a method for identifying karst goaf areas based on multi-source data fusion. Background Technology

[0002] In projects such as mountain highways, river valley bridges, and the renovation and expansion of old mining areas, bridge pile foundations often traverse soluble rock strata such as limestone and dolomite, and may be superimposed on former coal mines, gypsum mines, bauxite mines, and other mining subsidence areas. Hidden karst caves, filled karst caves, dissolution fracture zones, mining subsidence zones, tunnels, and water-rich fracture zones within the bearing capacity of the pile foundation will weaken the continuity of the bearing stratum at the pile tip, alter the distribution of pile tip reaction force and pile side friction, and affect the stress coordination of adjacent piles under the same pile cap.

[0003] Current methods for identifying karst goaf areas typically employ geological surveys, drilling, geophysical exploration, remote sensing deformation data, hydrogeological data, and mining data to delineate anomalous areas and classify their risks. Geophysical exploration can reveal anomalous characteristics such as resistivity, wave velocity, electromagnetic response, or radar reflection; drilling can expose local strata and filling conditions; mining data can reflect the mining scope and strata; and remote sensing deformation can reflect signs of surface subsidence or deformation.

[0004] However, within the influence range of bridge pile foundations, anomalous responses do not only correspond to the presence of cavities. Thick overburden layers, water-rich clay layers, fractured zones, filled karst caves, water-rich fissures, and mining-induced fracture zones can all generate similar low-resistivity, low-velocity, strong or weak reflection responses. After roof collapse, fissure expansion, groundwater recharge, and compaction, the anomalous characteristics of old mining areas overlap with those of natural karst infill bodies and water-rich fractured zones. When making judgments based solely on the delineation of a single anomaly, it is easy to confuse karst, mining-induced, and hydrological disturbance causes, leading to switching between different causal interpretations for supplementary investigations, pile location avoidance, grouting reinforcement, and pile length adjustments.

[0005] Furthermore, bridge pile foundations have clearly defined piers, abutments, pile groups, pile end bearing strata, and the associated range of adjacent piles. However, traditional regional karst goaf identification often relies on survey lines, boreholes, grids, or geological units, with results typically manifesting as planar anomaly zones, profile anomaly segments, or regional risk zoning. Such results are difficult to directly correlate with the causal category and risk status within the influence range of a single bridge pile foundation, and also fail to reflect the risk transmission relationship between adjacent piles caused by disturbance of the same anomaly or bearing stratum.

[0006] When inconsistencies arise between geophysical anomalies, borehole revelation, mining boundaries, remote sensing deformation, hydrological anomalies, and construction anomalies, existing methods often rely on intensified drilling, joint re-surveying, and manual interpretation. This approach struggles to simultaneously preserve conflicting causal evidence within the same bridge pile foundation's influence area, and it is difficult to distinguish the correspondence between shallow disturbances, deep karst, mining-induced crack zones, and hydrological disturbances. Consequently, the identification results lack data support corresponding to the causal source, conflicting evidence, and the pile foundation's treatment target.

[0007] Therefore, when there are conflicting pieces of evidence from multiple sources within the influence range of bridge pile foundations, it is difficult to separate the causes of karst formation, mining subsidence, and hydrological disturbance, and the identification results are difficult to correspond to the technical issues of single piles, pile groups, and piers. Summary of the Invention

[0008] This application provides a method for identifying karst goaf areas based on multi-source data fusion, which solves the problem of difficulty in separating the causes of multiple anomalies within the influence range of bridge pile foundations.

[0009] This invention provides a method for identifying karst goaf areas based on multi-source data fusion, comprising:

[0010] The bridge pile foundation influence body is generated based on bridge pier data, abutment data, pile foundation design data, pile end bearing layer data, and adjacent pile relationship data. The bridge pile foundation influence body is a three-dimensional engineering identification unit covering the pile body, pile end bearing layer, and the associated range of adjacent piles.

[0011] Acquire multi-source geological data corresponding to the bridge pile foundation influence body and register it to an engineering identification coordinate system indexed by route mileage, lateral offset, and stratum depth;

[0012] Genetic evidence records, including karst genesis evidence, mining subsidence evidence, hydrological disturbance evidence, and shallow disturbance evidence, are generated within each bridge pile foundation impact area. Data source applicability identifiers are configured to distinguish between valid, restricted, conflicting, and pending verification states based on the overburden disturbance state, depth-layer correspondence, spatial overlap, acquisition time, and data accuracy level.

[0013] Based on shallow interference evidence, geophysical anomaly data are marked for interference. The retained deep anomalies are matched with karst genesis evidence, mining subsidence evidence, and hydrological disturbance evidence to generate genetic candidate results.

[0014] When multiple candidate causal results correspond to the same bridge pile foundation affected body and the causal categories are mutually exclusive, an evidence conflict type is generated and the corresponding candidate causal results are retained.

[0015] Output the karst goaf identification results of the bridge pile foundation affected body. The karst goaf identification results include the causal category, risk level, evidence source identifier, evidence conflict type, and supplementary investigation priority.

[0016] In some embodiments, the multi-source geological data includes geophysical anomaly data, and at least two of the following: engineering geological survey data, borehole exposure data, mining data, remote sensing deformation data, hydrogeological data, and construction process data.

[0017] When each type of multi-source geological data enters the engineering identification coordinate system, it generates an evidence source identifier, a collection time identifier, and a spatial accuracy identifier.

[0018] In some embodiments, generating the bridge pile foundation influence body includes:

[0019] A single pile influence body is generated based on a single pile, a pile group influence body is generated based on multiple piles under the same pier cap, and a pier influence body is generated based on the planar projection relationship of adjacent pier caps and the continuous range of the bearing stratum at the pile tip. Each bridge pile foundation influence body records the pier number, pile number, pile tip elevation, bearing stratum range, and adjacent pile number.

[0020] In some embodiments, registering the multi-source geological data to an engineering identification coordinate system includes:

[0021] The data obtained from borehole exposure are converted into exposure strata, exposure objects, and exposure status; geophysical anomaly data are converted into anomaly spatial range, anomaly depth strata, and anomaly response type; mining data are converted into mining range, mining strata, and mining boundaries; remote sensing deformation data are converted into deformation range and deformation continuity status; and hydrogeological data are converted into groundwater activity strata and water-rich status.

[0022] In some embodiments, the evidence of karst formation includes at least one of the following: distribution of soluble rocks, karst landforms, exposure of caves, exposure of dissolution fracture zones, and groundwater connectivity.

[0023] The evidence for the formation of the goaf includes at least one of the following: mining area, mining strata, goaf boundary, rift zone, and historical subsidence;

[0024] The evidence of hydrological disturbance includes at least one of groundwater level changes, water abundance anomalies, water inrush records, and grout leakage records.

[0025] The shallow interference evidence includes at least one of pipelines, manhole covers, culverts, bridge abutment structures, road surface voids, loose backfill material, and metal reflection interference.

[0026] In some embodiments, generating candidate causal results includes:

[0027] When a deep anomaly corresponds to at least one of the following: distribution of soluble rocks, karst landforms, exposure of caves, exposure of dissolution fracture zones, and groundwater connectivity, a karst genesis candidate result is generated.

[0028] When a deep anomaly corresponds to at least one of the following: mining area, mining stratum, goaf boundary, caving zone, and historical subsidence, a goaf genesis candidate result is generated.

[0029] When a deep anomaly corresponds to at least one of groundwater level changes, water-rich anomalies, water inrush records, and grout leakage records, a candidate hydrological disturbance result is generated.

[0030] In some embodiments, the types of evidence conflict generated include:

[0031] When geophysical anomaly data and borehole exposure data are inconsistent in at least one of the object type, exposure status and depth stratum within the same bridge pile foundation influence body, an exploration-exposure conflict type is generated.

[0032] When the deformation range shown by remote sensing deformation data is inconsistent with the spatial range of deep anomalies, a deformation response conflict type is generated.

[0033] When mining data and karst genesis evidence are simultaneously associated with the same bridge pile foundation influence body, and both have a corresponding relationship with at least one of the influence ranges of the same deep anomaly and the same pile end bearing layer, a composite genetic conflict type is generated.

[0034] When shallow interference evidence overlaps with deep anomalies within a planar range, and the influence range of shallow interference evidence corresponds to the interpretability range of deep anomalies, an interference interpretation conflict type is generated.

[0035] When at least one of the abnormal layers, abnormal objects, and abnormal states shown in the construction process data is inconsistent with the existing candidate results for causes, a construction feedback conflict type is generated.

[0036] In some embodiments, after generating an evidence conflict type, an evidence conflict record is generated, which retains the candidate results of the causes of the conflict, the evidence source identifier, the triggering evidence source identifier, the conflict occurrence layer, and the bridge pile foundation influence body number.

[0037] When there is a conflict type of exploration and exposure, the evidence conflict record retains the anomaly response type corresponding to the geophysical anomaly data, the exposure object and exposure status corresponding to the borehole exposure data; when there is a conflict type of deformation response, the evidence conflict record retains the deformation range, deformation continuation status and deep anomaly range corresponding to the remote sensing deformation data.

[0038] When there is a complex genetic conflict type, the evidence conflict record simultaneously retains the karst genetic candidate results and the goaf genetic candidate results, and records the evidence source identifiers corresponding to the mining data, karst genetic evidence and geophysical anomaly data in the evidence conflict record;

[0039] When there are conflicting interpretation types, the evidence conflict record retains shallow interfering evidence, deep anomalies and corresponding interference categories, and writes the deep anomalies as retained objects into the evidence conflict record;

[0040] When there is a construction feedback conflict type, the construction anomaly shown in the construction process data is converted into construction feedback evidence, and the construction feedback evidence is associated with the bridge pile foundation affected by the construction anomaly.

[0041] When a construction anomaly is spatially or hierarchically related to the impact body of adjacent bridge pile foundations determined based on adjacent pile relationship data, the construction feedback evidence will be synchronously associated with the impact body of the adjacent bridge pile foundations.

[0042] In some embodiments, the risk levels include high risk, medium risk, and low risk;

[0043] Candidate causal results intersecting with the pile tip bearing stratum are identified as high-risk; candidate causal results located within the lateral influence range of the pile body and intersecting with the associated range of adjacent piles are identified as medium-risk; and candidate causal results located at the edge of the bridge pile foundation influence body and separated from the pile tip bearing stratum are identified as low-risk.

[0044] Data coverage gaps are gaps in multi-source geological data within the influence body of bridge pile foundations, resulting from at least one of the following: missing spatial range, missing depth and stratum, missing acquisition time, missing spatial accuracy identification, missing evidence source identification, and missing coordinate registration results. Supplementary exploration priorities are generated according to risk level, type of evidence conflict, and data coverage gaps.

[0045] In some embodiments, the construction process data includes at least one of drilling status data, grout leakage records, drill bit loss records, borehole collapse records, water inrush records, mud level change records, drill cuttings exposure records, and actual rock entry status records.

[0046] When construction process data indicates that there is a construction anomaly in a certain bridge pile foundation affected body, the construction anomaly is converted into construction feedback evidence, and the karst mining area identification results, evidence conflict type, and supplementary investigation priority of the bridge pile foundation affected body and the adjacent bridge pile foundation affected bodies determined based on adjacent pile relationship data are updated.

[0047] Through the above technical solution, the present invention can achieve at least the following beneficial effects:

[0048] This invention uses the influence body of bridge pile foundations as the identification object for karst mining subsidence areas. It incorporates bridge piers, abutments, pile bodies, pile end bearing layers, and the associated range of adjacent piles into the same three-dimensional engineering identification unit, establishing a correspondence between the abnormal locations and abnormal strata of multi-source geological data and the engineering objects. By generating causal evidence records within each bridge pile foundation influence body and recording karst causal evidence, mining subsidence causal evidence, hydrological disturbance evidence, and shallow disturbance evidence separately, it is possible to distinguish different anomaly sources within the same pile foundation influence area. This allows geophysical anomalies, borehole exposure, mining data, remote sensing deformation, hydrological information, and construction process information to be mapped to specific causal categories, reducing the switching of treatment criteria caused by the confusion of karst, mining subsidence, and hydrological disturbance causes.

[0049] By configuring data source applicability identifiers based on overburden interference status, depth-stratum correspondence, spatial overlap, acquisition time, and data accuracy level, data that can be directly used for identification, data affected by overburden, data with conflicting evidence, and data lacking necessary fields can be processed separately. This data processing method incorporates factors such as thick overburden, water-rich clay layers, shallow structures, and data time lag into the identification process, preventing single geophysical anomalies from being directly used as the basis for causation judgment, and maintaining a correspondence between candidate causal results and the available data status.

[0050] By marking geophysical anomaly data with interference based on shallow interference evidence, and then matching the retained deep anomalies with evidence of karst genesis, goaf genesis, and hydrological disturbance, it is possible to distinguish shallow pipelines, manhole covers, culverts, bridge abutment structures, pavement voids, loose backfill material, and metal reflection interference from the deep anomaly interpretation chain. After deep anomalies are included in the genetic matching, karst cave exposure, soluble rock distribution, mining strata, goaf boundaries, water-rich anomalies, and grout leakage records correspond to different genetic candidate results, allowing filled karst caves, goaf fracture zones, and water-rich fracture zones to retain their respective genetic orientations even under similar anomaly responses.

[0051] When multiple candidate causal results correspond to the same bridge pile foundation impact body and there are conflicting evidence categories, an evidence conflict type is generated, and the corresponding candidate causal results are retained. This allows for the retention of conflict sources and candidate causal results when there are inconsistencies between geophysical anomalies and borehole findings, overlap between mining data and karst evidence, overlap between shallow interference and deep anomalies, and inconsistencies in construction anomaly feedback. The output karst goaf identification results also include causal category, risk level, evidence source identifier, evidence conflict type, and supplementary investigation priority, enabling the identification results to correspond to risk assessment and supplementary investigation arrangements within the scope of single piles, pile groups, and piers. Attached Figure Description

[0052] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments will be briefly described below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation on the scope of this application.

[0053] Figure 1 This is a flowchart of the karst goaf identification method based on multi-source data fusion in the embodiments. Detailed Implementation

[0054] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0055] All terms used in this application (including technical and scientific terms) have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein should be interpreted in a manner consistent with the context of this specification, and not in an idealized or overly rigid way.

[0056] The relevant terms and concepts involved in the embodiments of this application are as follows:

[0057] A bridge pile foundation influence body refers to a three-dimensional engineering identification unit generated using bridge pier data, abutment data, pile foundation design data, pile bearing layer data, and adjacent pile relationship data as constraints. The bridge pile foundation influence body records the pier number, abutment number, pile number, pile coordinates, pile diameter, design pile length, pile end elevation, pile bearing layer name, pile bearing layer range, adjacent pile numbers, and adjacent abutment numbers. The influence body number is a data index field used to uniquely identify the bridge pile foundation influence body. Single pile influence bodies, pile group influence bodies, and pier influence bodies are each configured with independent influence body numbers, and their relationships are established through pier number, abutment number, pile number, adjacent pile number, and adjacent abutment number. The current identification time refers to the batch time written when the bridge pile foundation influence body performs identification calculations. An identification batch refers to a data processing batch that performs identification calculations on bridge pile foundation influence bodies within a bridge construction site at the same current identification time. The identification batch records the batch number, the current identification time, and the set of influence body numbers involved in the calculation.

[0058] The engineering identification coordinate system refers to a spatial data organization system indexed by route mileage, lateral offset, and stratigraphic depth. Route mileage represents the position of the target bridge work site along the route direction; lateral offset represents the lateral position of the target bridge work site relative to the route centerline; and stratigraphic depth represents the vertical position of the target bridge work site relative to a unified elevation datum. The unified elevation datum refers to the same vertical benchmark used within the bridge work site to express stratigraphic depth, borehole depth, geophysical anomaly depth, mining strata, and groundwater activity strata. Before multi-source geological data enters the engineering identification coordinate system, the plane coordinates are converted to route mileage and lateral offset, and the original elevation, borehole depth, interpretation depth, and mining strata are converted to stratigraphic depth relative to the unified elevation datum. Data lacking unified elevation datum conversion relationships constitutes a coordinate registration gap and enters a pending verification state.

[0059] Multi-source geological data refers to engineering geological survey data, borehole exposure data, geophysical anomaly data, mining data, remote sensing deformation data, hydrogeological data, and construction process data that are spatially related to the bridge pile foundation impact body. Evidence source identification refers to the source tracing field that records the source category, source file number, acquisition time identifier, and data formation stage of the multi-source geological data. Each type of multi-source geological data records the evidence source identifier, acquisition time identifier, spatial accuracy identifier, data coverage, depth horizon, and data status. Before multi-source geological data enters the genetic evidence recording stage, necessary fields include the evidence source identifier, acquisition time identifier, spatial accuracy identifier, spatial range field, and depth horizon field; the spatial range field records the planar location, boundary range, and coverage integrity status, while the depth horizon field records the depth benchmark, upper depth limit, lower depth limit, and center depth.

[0060] Causal evidence records refer to structured evidence data formed within the affected body of bridge pile foundations. These records include the affected body number, evidence category, evidence source identifier, evidence spatial range, evidence depth stratigraphy, exposed object, exposure status, collection time identifier, spatial accuracy identifier, data source applicability identifier, candidate causal results, and evidence conflict type. The evidence category refers to one of the following: karst causal evidence, mining subsidence causal evidence, hydrological disturbance evidence, or shallow disturbance evidence. Causal evidence records are stored within the same identification batch, indexed by the affected body number. When the same multi-source geological data covers multiple affected bodies of bridge pile foundations, causal evidence records corresponding to each affected body number are generated separately, and the same evidence source identifier is retained in each record.

[0061] Data source applicability identifiers refer to data identifiers characterizing the status of multi-source geological data in the current bridge pile foundation influence body for causal identification. Data source applicability identifiers include valid status, restricted status, conflict status, and pending verification status. Data accuracy level refers to the level data converted from spatial accuracy identifier, positioning accuracy category, depth positioning accuracy category, and interpretation accuracy category. Spatial accuracy identifier is the original accuracy field generated when entering the engineering identification coordinate system, and data accuracy level is the evaluation field generated for participating data source applicability identifiers. One data source applicability identifier is configured for the same causal evidence record within the same identification batch. When the same causal evidence record simultaneously meets multiple status conditions, the data source applicability identifier is determined in the order of conflict status, pending verification status, restricted status, and valid status. Mutually exclusive causal categories refer to situations where causal evidence records within the same bridge pile foundation influence body point to different causal categories at the same depth layer or within the same deep anomaly range, and the corresponding exposure objects, spatial ranges, construction feedback status, or mining boundary statuses for each causal category cannot be simultaneously established. When mining data and karst genetic evidence both point to the same deep anomaly and have spatial and stratigraphic correspondences, they should be treated as a composite genetic conflict type.

[0062] Explained shallow anomalies refer to the anomalous portions of geophysical anomaly data that have a spatial correspondence with shallow interference evidence and are marked as sources of shallow interference. Deep anomalies refer to the anomalous portions of geophysical anomaly data that are retained after being marked for shallow interference and included in the generation of causal candidate results. The influence range of shallow engineering structures refers to the spatial influence range corresponding to at least one type of shallow interference evidence, such as pipelines, manhole covers, culverts, bridge abutment structures, pavement voids, loose backfill material, and metal reflection interference. Causal candidate results refer to the candidate identification results formed by matching deep anomalies with causal evidence records. Causal candidate results include karst causal candidate results, mining subsidence causal candidate results, and hydrological disturbance candidate results. Candidate result status refers to the data field that records the processing status of causal candidate results in the current identification batch. The candidate result status includes generated status, conflict retention status, and feedback update status. The generated status indicates that the deep anomaly has been matched with the corresponding causal evidence. The conflict retention status indicates that the causal candidate result has participated in the evidence conflict record and has been retained. The feedback update status indicates that the causal candidate result has been updated after the construction feedback evidence has been entered.

[0063] The karst goaf identification results refer to the output data generated after the causal separation and evidence conflict resolution of the impact bodies of bridge pile foundations. The karst goaf identification results include the impact body number, causal category, risk level, evidence source identifier, evidence conflict type, and priority of supplementary investigation.

[0064] Example 1:

[0065] like Figure 1As shown, this embodiment provides a method for identifying karst goaf areas oriented towards the influence body of bridge pile foundations. The method first generates the influence body of bridge pile foundations and registers multi-source geological data to the engineering identification coordinate system; then, it generates causal evidence records and data source applicability identifiers within the influence body of bridge pile foundations; subsequently, it distinguishes between shallow disturbances and deep anomalies, and generates causal candidate results based on karst causal evidence, goaf causal evidence, and hydrological disturbance evidence; when there are conflicting evidence, the corresponding causal candidate results are retained, and the karst goaf area identification result is output. This embodiment includes the following steps:

[0066] Step S1: Generate a bridge pile foundation influence body based on bridge pier data, abutment data, pile foundation design data, pile end bearing layer data, and adjacent pile relationship data. The bridge pile foundation influence body is a three-dimensional engineering identification unit covering the pile body, pile end bearing layer, and the associated range of adjacent piles.

[0067] Step S2: Obtain multi-source geological data corresponding to the bridge pile foundation influence body and register it to the engineering identification coordinate system indexed by route mileage, lateral offset and stratum depth;

[0068] Step S3: Generate causal evidence records within each bridge pile foundation impact area. Causal evidence records include karst causal evidence, mining subsidence causal evidence, hydrological disturbance evidence, and shallow disturbance evidence. Configure data source applicability identifiers for causal evidence records.

[0069] Step S4: The data source applicability identifier is generated according to the overlay interference status, depth layer correspondence, spatial overlap, acquisition time relationship, and data accuracy level. When the causal evidence record meets the requirements of spatial coverage, layer correspondence, and accuracy, the data source applicability identifier is configured as valid. When the causal evidence record is affected by at least one of overlay interference, incomplete spatial coverage, and data time lag, and can be registered with complete necessary fields, the data source applicability identifier is configured as restricted. When different causal evidence records point to mutually exclusive causal categories within the same bridge pile foundation influence body, the data source applicability identifier is configured as conflicting. When the causal evidence record has at least one of the following conditions: missing necessary fields and inability to register spatial location, the data source applicability identifier is configured as pending verification.

[0070] Step S5: Based on shallow interference evidence, the geophysical anomaly data is marked for interference. The deep anomalies retained after interference marking are matched with karst genesis evidence, mining subsidence genesis evidence, and hydrological disturbance evidence respectively to generate genetic candidate results.

[0071] Step S6: When multiple candidate causal results correspond to the same bridge pile foundation influence body and the causal categories are mutually exclusive, generate an evidence conflict type and retain the corresponding candidate causal results.

[0072] Step S7: Output the karst goaf identification results of the bridge pile foundation impact body. The karst goaf identification results include the causal category, risk level, evidence source identifier, evidence conflict type, and supplementary investigation priority.

[0073] In one implementation, when shallow interference evidence and geophysical anomaly data overlap in planar plane and the depth layer is within the influence range of the shallow engineering structure, the corresponding geophysical anomaly data is marked as interpreted shallow anomaly. When geophysical anomaly data and shallow interference evidence overlap in planar plane but the depth layer extends towards the pile end bearing layer, the shallow overlapping portion is marked as interpreted shallow anomaly, and the portion extending towards the pile end bearing layer is retained as deep anomaly.

[0074] When matching deep anomalies with evidence of karst origin, mining subsidence, and hydrological disturbance, the matching criteria include spatial overlap, depth-stratum correspondence, stratigraphic type correspondence, evidence source identifier, and collection time identifier. When deep anomalies correspond to the distribution of soluble rocks, karst landforms, cave exposure, dissolution fracture zones, or groundwater connectivity, karst origin candidate results are formed. When deep anomalies correspond to mining areas, mining strata, mining subsidence boundaries, fracture zones, or historical subsidence, mining subsidence candidate results are formed. When deep anomalies correspond to groundwater level changes, water-rich anomalies, water inrush records, or grout leakage records, hydrological disturbance candidate results are formed. Spatial overlap, depth-stratum correspondence, and stratigraphic type correspondence are all determined within the engineering identification coordinate system. Spatial overlap is determined based on the overlap between the planar projection range of the deep anomaly and the planar range of the genetic evidence record. Depth stratigraphic correspondence is determined based on the overlap or deviation between the depth stratigraphic position of the deep anomaly and the depth stratigraphic position of the genetic evidence record. Stratigraphic type correspondence is determined based on the correspondence between the stratigraphic position of the deep anomaly and the soluble rock stratigraphic position, ore-bearing stratigraphic position, or aquifer stratigraphic position. If any of the above relationships lacks a calculable boundary or depth benchmark, the corresponding genetic evidence record enters a pending verification state.

[0075] In one implementation, the overburden interference state is determined based on the overburden thickness category, overburden water content, shallow engineering structure distribution, and shallow metal reflection interference, used to characterize the influence of the overburden and shallow interference factors on the geophysical anomaly interpretation results. The depth-stratum correspondence is determined based on the spatial relationship between the corresponding depth in the causal evidence record and the pile body range, pile tip bearing layer range, and adjacent pile associated range, used to characterize the correspondence between the data-pointed stratum and key engineering components within the bridge pile foundation's influence body. The spatial overlap relationship is determined based on the overlap between the planar range corresponding to the causal evidence record and the projected range of the bridge pile foundation's influence body. The acquisition time relationship is determined based on the chronological relationship between the acquisition time, construction time, and data formation time of the corresponding data in the causal evidence record. The data accuracy level is determined based on the source category, positioning accuracy category, and interpretation accuracy category in the evidence source identifier.

[0076] When the data spatial range corresponding to a certain causal evidence record effectively overlaps with the influence body of the bridge pile foundation, the depth layer corresponds to the range of the pile body, the range of the pile end bearing layer or the range associated with adjacent piles, the collection time relationship meets the requirements of the current identification batch, and the data accuracy level meets the current identification requirements, the data source applicability flag of the causal evidence record is configured as valid.

[0077] When the data corresponding to a causal evidence record is affected by overburden interference, shallow engineering structure interference, shallow metal reflection interference, incomplete spatial coverage, delayed acquisition time, or insufficient data accuracy, but coordinate registration can still be completed and necessary fields retained, the data source applicability flag of that causal evidence record is configured as restricted. Causal evidence records in a restricted state are not directly used as the basis for determining the causal category alone, but can participate in the generation of causal candidate results together with other valid or restricted causal evidence records.

[0078] When different causal evidence records point to different causal categories within the same bridge pile foundation influence area, and form a mutually exclusive relationship in terms of object type, depth layer, spatial range, or construction feedback status, the data source applicability identifier of the relevant causal evidence records will be configured as conflicting. Causal evidence records in conflicting status will not be deleted, but will be sent to evidence conflict processing to generate the corresponding evidence conflict type.

[0079] When a causal evidence record lacks an evidence source identifier, collection time identifier, spatial accuracy identifier, boundary accuracy information, depth benchmark information, or coordinate registration results, or when spatial location registration cannot be completed, the data source applicability identifier of that causal evidence record is configured to a pending verification status. Causal evidence records in the pending verification status are used to indicate data coverage gaps and are not used as a basis for independently determining the causal category.

[0080] In one implementation, the multi-source geological data includes at least three of the following: engineering geological survey data, borehole exposure data, geophysical anomaly data, mining data, remote sensing deformation data, hydrogeological data, and construction process data.

[0081] When each type of multi-source geological data enters the engineering identification coordinate system, it generates an evidence source identifier, a collection time identifier, and a spatial accuracy identifier.

[0082] The evidence source identifier records the source category and source document number of the multi-source geological data. Source categories include engineering geological surveys, borehole exposure, geophysical exploration, mining data, remote sensing deformation, hydrogeology, and construction processes. The acquisition time identifier records the acquisition date, data formation date, or construction record date corresponding to the multi-source geological data. The spatial accuracy identifier records the horizontal positioning accuracy category, depth positioning accuracy category, and coverage integrity status of the multi-source geological data.

[0083] When multi-source geological data lacks any of the necessary fields such as planar location, depth horizon, or acquisition time, the data source applicability flag for the corresponding data is configured to be pending verification. When multi-source geological data covers the influence body of bridge pile foundations but the spatial accuracy flag is lower than the engineering identification requirements, the data source applicability flag for the corresponding data is configured to be restricted. When multi-source geological data from different sources point to mutually exclusive causal categories within the same bridge pile foundation influence body, the data source applicability flag for the corresponding data is configured to be conflicting.

[0084] In one implementation, generating the bridge pile foundation influence body includes:

[0085] A single pile influence body is generated based on a single pile, a pile group influence body is generated based on multiple piles under the same pile cap, and a pier influence body is generated based on the planar projection relationship of adjacent pile caps and the continuous range of the bearing stratum at the pile tip; each bridge pile foundation influence body records the pier number, pile number, pile tip elevation, bearing stratum range, and adjacent pile number;

[0086] The influence body of a single pile is generated based on the spatial range of the pile body, the bearing stratum at the pile tip, and the influence range around the pile. The influence body of a pile group is generated based on the union of the influence bodies of all single piles under the same pile cap, and the adjacent pile numbers between pile numbers under the same pile cap are recorded. The influence body of a pier / abutment is generated based on the plane range of the pile cap corresponding to the same pier / abutment, the range of the influence body of the pile group, and the continuous range of the bearing stratum at the pile tip, and the adjacent pile cap numbers are recorded.

[0087] When two bridge pile foundation impact bodies spatially overlap, the overlapping area is recorded separately for single pile impact body, pile group impact body, and pier impact body. The causal evidence records formed within the overlapping area are simultaneously linked to the corresponding single pile impact body, pile group impact body, and pier impact body, and their respective impact body numbers are retained.

[0088] In one implementation, registering multi-source geological data to an engineering identification coordinate system includes:

[0089] The data is converted from borehole exposure data into exposure strata, exposure objects and exposure status; geophysical anomaly data into anomaly spatial range, anomaly depth strata and anomaly response type; mining data into mining range, mining strata and mining boundary; remote sensing deformation data into deformation range and deformation continuity status; and hydrogeological data into groundwater activity strata and water-rich status.

[0090] When borehole exposure data is entered into the engineering identification coordinate system, the exposed strata are generated based on the borehole opening location, borehole trajectory, exposure depth, and unified elevation datum. When geophysical anomaly data is entered into the engineering identification coordinate system, the anomaly spatial range is generated based on the survey line location, anomaly interpretation depth, anomaly spatial extent, and anomaly response type. When mining data is entered into the engineering identification coordinate system, the mining area is generated based on the mining map coordinates, mining strata, mining boundaries, and mining time. When remote sensing deformation data is entered into the engineering identification coordinate system, the deformation range is generated based on the image coverage, deformation zone boundaries, and deformation continuity. When hydrogeological data is entered into the engineering identification coordinate system, groundwater activity strata are generated based on the observation location, water level change records, aquifers, and water-rich status.

[0091] The registered multi-source geological data are all associated with the corresponding bridge pile foundation impact bodies. The association fields include impact body number, evidence source identifier, spatial overlap status, depth and layer correspondence status, acquisition time identifier, and spatial accuracy identifier.

[0092] In one embodiment, evidence of karst formation includes at least one of the following: distribution of soluble rocks, karst landforms, exposure of caves, exposure of dissolution fracture zones, and groundwater connectivity.

[0093] Evidence for the formation of goaf includes at least one of the following: mining extent, mining stratigraphic position, goaf boundary, rift zone, and historical subsidence;

[0094] Evidence of hydrological disturbance includes at least one of the following: groundwater level changes, water abundance anomalies, water inrush records, and grout leakage records;

[0095] Shallow interference evidence includes at least one of the following: pipelines, manhole covers, culverts, bridge abutment structures, pavement voids, loose backfill material, and metal reflection interference;

[0096] Essential fields for karst formation evidence include soluble rock strata, karst development location, cave exposure status, dissolution fracture zone exposure status, groundwater connectivity status, and corresponding evidence source identifiers. Essential fields for mining-induced subsidence evidence include mining area, mining strata, subsidence boundary, location of fracture zones, historical subsidence area, and corresponding evidence source identifiers. Essential fields for hydrological disturbance evidence include groundwater activity strata, groundwater level changes, water abundance status, water inrush records, grout leakage records, and corresponding evidence source identifiers. Essential fields for shallow disturbance evidence include the type of disturbed object, the location of the disturbed object, the depth of the disturbed object, the extent of the disturbance's impact, and corresponding evidence source identifiers.

[0097] Evidence for karst formation, mining subsidence, hydrological disturbance, and shallow disturbance is stored in the form of causal evidence records. Each causal evidence record is associated with one bridge pile foundation affected body; when the same evidence covers multiple bridge pile foundation affected bodies simultaneously, causal evidence records corresponding to each bridge pile foundation affected body are generated separately.

[0098] In one implementation, generating candidate causal results includes:

[0099] When a deep anomaly corresponds to at least one of the following: distribution of soluble rocks, karst landforms, exposure of caves, exposure of dissolution fracture zones, and groundwater connectivity, a karst genesis candidate result is generated.

[0100] When a deep anomaly corresponds to at least one of the following: mining area, mining stratum, goaf boundary, caving zone, and historical subsidence, a goaf genesis candidate result is generated.

[0101] When a deep anomaly corresponds to at least one of groundwater level changes, water-rich anomalies, water inrush records, and slurry leakage records, a candidate hydrological disturbance result is generated.

[0102] The candidate causal results are stored in the form of candidate result records. The candidate result record includes the influencing body number, candidate causal category, corresponding deep anomaly, matching evidence category, evidence source identifier of the matching evidence, spatial overlap status, depth layer correspondence status, data source applicability identifier, and candidate result status.

[0103] When the same deep anomaly matches both karst origin evidence and hydrological disturbance evidence, the candidate result record includes both karst origin candidate results and hydrological disturbance candidate results. When the same deep anomaly matches both mining subsidence origin evidence and hydrological disturbance evidence, the candidate result record includes both mining subsidence origin candidate results and hydrological disturbance candidate results. When the same deep anomaly matches both karst origin evidence and mining subsidence origin evidence, the candidate result record includes both karst origin candidate results and mining subsidence candidate results, and this candidate result record is then submitted to the evidence conflict resolution process.

[0104] In one implementation, the risk levels include high risk, medium risk, and low risk;

[0105] Candidate causal results intersecting with the pile tip bearing stratum are identified as high-risk; candidate causal results located within the lateral influence range of the pile body and intersecting with the associated range of adjacent piles are identified as medium-risk; and candidate causal results located at the edge of the bridge pile foundation influence body and separated from the pile tip bearing stratum are identified as low-risk.

[0106] Supplementary investigation priorities are generated based on risk level, type of evidence conflict, and data coverage gaps;

[0107] Risk levels are stored in the form of risk level records. These records include the affected body number, causal category, candidate result location, intersection status of the pile tip bearing layer, lateral impact status of the pile body, association status of adjacent piles, type of evidence conflict, and data coverage gap. A data coverage gap refers to a gap record formed by at least one of the following conditions within the affected body of the bridge pile foundation: missing spatial extent, missing depth / stratum, missing acquisition time, missing spatial accuracy identifier, missing evidence source identifier, and missing coordinate registration results. The data coverage gap record includes the affected body number, missing data category, missing field name, missing spatial extent, missing depth / stratum, corresponding evidence source identifier, and items to be verified. When generating supplementary exploration priorities, data coverage gaps, risk level records, and evidence conflict records are all included in the ranking process.

[0108] Supplementary investigation priorities are generated based on risk level records. When a candidate causal result intersects with the pile tip bearing stratum and there is an evidence conflict type, the supplementary investigation priority is configured as the first priority. When a candidate causal result is located within the lateral influence range of the pile body and is associated with adjacent pile numbers, the supplementary investigation priority is configured as the second priority. When a candidate causal result is located at the edge of the bridge pile foundation influence body and there is a data coverage gap, the supplementary investigation priority is configured as the third priority. The output of the supplementary investigation priority includes the influence body number, priority category, causal category to be verified, depth layer to be verified, data source to be verified, and corresponding evidence conflict type.

[0109] In one embodiment, the construction process data includes at least one of drilling status data, grout leakage records, drill bit loss records, borehole collapse records, water inrush records, mud level change records, drill cuttings exposure records, and actual rock entry status records.

[0110] When construction process data indicates that there is a construction anomaly in a certain bridge pile foundation affected body, the construction anomaly is converted into construction feedback evidence, and the karst mining area identification results, evidence conflict type, and supplementary investigation priority of the bridge pile foundation affected body and the adjacent bridge pile foundation affected bodies determined based on adjacent pile relationship data are updated.

[0111] Construction feedback evidence is stored in the form of construction anomaly records. Construction anomaly records include construction pile number, time of anomaly occurrence, depth of anomaly occurrence, drilling status, grout leakage status, drill bit loss status, borehole collapse status, water inrush status, mud level change status, drill cuttings exposure status, actual rock penetration status, and construction record source identifier.

[0112] After a construction anomaly record is generated, it is associated with the corresponding bridge pile foundation affected body based on the construction pile number, and further associated with the affected body of the adjacent bridge pile foundation based on the adjacent pile number. Once construction feedback evidence is entered into the causal evidence record, the hydrological disturbance evidence, karst causal evidence, or mining subsidence causal evidence for the corresponding bridge pile foundation affected body is updated, and the candidate result record, evidence conflict type, risk level record, and supplementary investigation priority are updated simultaneously. The updated karst mining subsidence area identification result record includes the update batch, update time, triggering construction anomaly record, and previous result identifier.

[0113] In a preferred embodiment of Example 1, when generating the data source applicability identifier, any causal evidence record within a single bridge pile foundation influence body is used as the evaluation object. The overburden interference state, depth-layer correspondence, spatial overlap, acquisition time relationship, and data accuracy level are converted into applicability control quantities within the same value range. The applicability control quantities include overburden interference quantity, depth-layer correspondence quantity, spatial overlap quantity, time validity quantity, and data accuracy reliability quantity. The comprehensive applicability value of the causal evidence record is then generated using the following formula:

[0114] ,

[0115] in, For the first The overall applicability score of each causal evidence record; For the first The weight of each causal evidence record corresponding to the interference state of the overlay layer; For the first The amount of interference from the overlay of causal evidence records; For the first The weight of the corresponding depth layer relationship in the causal evidence records; For the first The depth of the stratigraphic sequence corresponding to the causal evidence record; For the first The weight of spatial overlap relationships corresponding to each piece of causal evidence record; For the first The amount of spatial overlap in the records of causal evidence; For the first The weight of the relationship between the collection time and the causal evidence records; For the first The effective time frame of the causal evidence record; For the first The weight of the precision level corresponding to each piece of causal evidence record; For the first The accuracy and reliability of the data recorded in the causal evidence section. , , , and The values ​​are all limited to 0~1, the weight values ​​are all not less than 0, and the sum of the weights corresponding to the same causal evidence record is 1.

[0116] The evidence conflict trigger value is a binary identifier used to record whether a causal evidence record triggers an evidence conflict type. The verification gap trigger value is a binary identifier used to record whether a causal evidence record has at least one verification gap, such as missing necessary fields, missing coordinate registration, missing depth benchmarks, or missing spatial accuracy identifiers. The applicability comprehensive value is used for ranking and determining the status of similar calculable evidence, but not for offsetting the failure of a single key condition. When the depth-stratum correspondence cannot be calculated, the spatial range lacks necessary boundaries, the acquisition time is missing, the spatial accuracy identifier is missing, or multi-source geological data cannot be registered to the engineering identification coordinate system, the verification gap trigger value of the corresponding causal evidence record is set to 1, and it is not marked as valid due to higher values ​​of other applicability control quantities. When the depth-stratum correspondence has reached a conclusion contrary to borehole exposure, construction feedback, or confirmed mined-out boundaries, the evidence conflict trigger value of the corresponding causal evidence record is set to 1, and it is not offset by increasing the weight or other control quantities. The overburden interference, depth-stratum correspondence, spatial overlap, time validity, and data accuracy reliability are all applicability control quantities within the range of 0 to 1. A larger value for the overburden interference indicates a stronger weakening effect of overburden and shallow layer interference on the interpretation of geophysical anomalies. Larger values ​​for depth-level correspondence, spatial overlap, temporal validity, and data accuracy reliability indicate higher degrees of correspondence, spatial overlap, temporal proximity, and accuracy reliability, respectively. All applicability control quantities are converted using a unified conversion method within the same identification batch before being included in the overall applicability value.

[0117] According to the The degree of overlap between the spatial extent of each causal evidence record and the affected area of ​​the bridge pile foundation, the extent of deep anomalies, or the target stratum is calculated. The higher the degree of overlap, the better. The larger the value, the less likely it is to be included when the corresponding spatial range lacks boundary precision. As a basis for determining the effective status, the verification gap trigger value of the corresponding cause evidence record is set to 1. According to the The conversion between the collection time of the causal evidence record and the current identification time is as follows: the shorter the time interval, the better. The larger the value, the less calculation is performed when the data collection time is missing. And set the verification gap trigger value of the corresponding causal evidence record to 1. According to the The conversion of data precision levels for each piece of causal evidence record: the higher the data precision level, the better. The larger the value, the less calculation is performed when the spatial precision identifier is missing or cannot be converted. And set the verification gap trigger value of the corresponding causal evidence record to 1.

[0118] The current identification time does not change with the system query time. Within the same batch, the effective time of causal evidence records is converted using the same current identification time; when re-identifying, the new batch time is rewritten and recorded in the evidence source identifier.

[0119] Weights can be calibrated based on bridge pile foundation impact body samples that have undergone borehole exposure and construction feedback verification. When the number of samples is insufficient, the weights corresponding to depth-layer correspondence and spatial overlap should not be lower than the weights corresponding to overburden interference state, acquisition time relationship, and data accuracy level. Calibration samples refer to bridge pile foundation impact body samples that have undergone borehole exposure, construction feedback verification, or supplementary geophysical verification, and whose causal category, spatial range, and depth-layer are definitive. Weights, overburden interference limitation thresholds, effective applicable thresholds, exploration conflict thresholds, deformation response conflict thresholds, composite causal conflict thresholds, and interference coverage thresholds can all be determined based on calibration sample quantiles, verification set integration, or industry limits. The calibration version number is a version field used to record weight values, threshold values, calibration sample range, and calibration time. Weights and thresholds under the same calibration version number use the same calibration sample range.

[0120] The overburden interference quantity is used to characterize the degree to which thick overburden, shallow structures, metallic reflectors, or loose backfill material weaken the interpretation of geophysical anomalies. Its calculation method is as follows:

[0121] ,

[0122] in, For the first The amount of interference from the overlay of causal evidence records; For the first The equivalent thickness of the covering layer corresponding to each causal evidence record; For the first The equivalent thickness of shallow interference corresponding to each causal evidence record; For the first The thickness is calculated by converting the anti-interference resolution capability of the corresponding data source within the target depth range for each genetic evidence record. The larger the value, the stronger the tolerance of the multi-source geological data corresponding to the genetic evidence record to interference from thick overburden and shallow layers. The spatial accuracy identifier, target depth, and applicable depth of the corresponding multi-source geological data are used for joint calibration; when When data is missing or cannot be converted from the source of evidence, a conservative approach is taken, using the smaller value corresponding to the lowest permissible capacity of similar multi-source geological data. and All values ​​are non-negative. Use values ​​greater than zero.

[0123] The equivalent thickness of the overburden is calculated from the upper and lower boundaries of the overburden in engineering geological survey data, borehole exposure data, and geophysical anomaly data. The equivalent thickness of shallow interference is calculated from the depth of the interfering object and the range of interference influence in shallow interference evidence. The target depth is the center depth of the deep anomaly, the center depth of the object to be verified, or the center depth of the pile end bearing layer. The target stratigraphic range is the depth range corresponding to the deep anomaly, the pile end bearing layer, the mining layer, or the groundwater activity layer.

[0124] When the target depth exceeds the applicable depth of the corresponding multi-source geological data, but the spatial extent of the anomaly can still be determined, the causal evidence record is not marked as valid, but rather as restricted; when When the depth cannot be calculated from the spatial precision identifier, target depth, and applicable depth of the corresponding multi-source geological data, the verification gap trigger value of the corresponding genetic evidence record is set to 1 and marked as pending verification.

[0125] The depth-stratum correspondence is used to characterize the degree of consistency between geophysical anomalies, borehole-revealed objects, mining strata, groundwater activity strata, and pile-end bearing strata or anomaly target strata. Its calculation method is as follows:

[0126] ,

[0127] in, For the first The depth of the stratigraphic sequence corresponding to the causal evidence record; For the first The central depth of the object to be verified in the causal evidence record; For the first The central depth of the reference object in the causal evidence record; For the first Each causal evidence record corresponds to the allowable depth deviation of the data source. The object to be verified can be a geophysical anomaly depth layer, a mining layer, or a groundwater activity layer, and the reference object can be a borehole-exposed layer, a pile-end bearing layer, or a confirmed mined-out boundary layer. The calibration is based on borehole depth error, geophysical inversion vertical resolution, and formation undulation amplitude; when A decrease indicates a weakening of the depth-stratum correspondence; when the difference between the borehole-revealed stratum and the geophysical anomaly depth exceeds... In such cases, the evidence of the cause shall be recorded as a source of conflict criterion.

[0128] and Before entering the calculation, the same depth reference is unified in the engineering identification coordinate system; when the object to be checked or the reference object is a depth range, the center depth of the depth range is used as the corresponding center depth. Values ​​greater than zero are used; no calculation is performed when the depth reference is inconsistent, the upper and lower boundaries of the depth interval are missing, or the allowable depth deviation cannot be calibrated. The corresponding causal evidence record will be marked as pending verification. If the depth benchmark has been unified and the depth deviation exceeds the allowable depth deviation, the corresponding causal evidence record will be used as the source of conflict judgment.

[0129] Based on the above calculations, the data source applicability identifier is generated according to the following formula:

[0130] hour, In a state of conflict,

[0131] hour, The status is pending verification.

[0132] and And in and When any one of them is established, It is in a restricted state.

[0133] and ,and and At the same time, It is in a valid state;

[0134] in, For the first Applicability identifier of the data source for each causal evidence record; For the first The evidence conflict trigger value for each causal evidence record; For the first The threshold for triggering a gap in the verification of causal evidence records; For the first The threshold for limiting interference in the overlay corresponding to each causal evidence record; For the first The effective applicable threshold corresponding to each causal evidence record. A value of 1 indicates that a conflict of evidence exists, while a value of 0 indicates that no conflict of evidence has been triggered. A value of 1 indicates the existence of missing boundary accuracy, missing acquisition time, missing coordinate transformation, or missing spatial accuracy level; a value of 0 indicates the absence of any of the above-mentioned verification gaps. The evidence conflict trigger value is generated from the discrimination results of the trigger conditions corresponding to exploration conflict, deformation response conflict, composite cause conflict, interference interpretation conflict, and construction feedback conflict; when any conflict trigger condition is met, Set to 1. The gap verification trigger value is generated from the evidence source identifier, acquisition time identifier, spatial accuracy identifier, coordinate registration result, depth benchmark consistency, and boundary accuracy status; if any necessary field is missing or cannot be converted, Set to 1.

[0135] After new borehole data, construction process data, or supplementary geophysical data are added to the engineering identification coordinate system, only the bridge pile foundations affected by the new data, as well as the bridge pile foundations with spatial or stratigraphic relationships among adjacent pile numbers, will be updated. Bridge pile foundations without spatial or stratigraphic relationships will not be updated. and .

[0136] and According to the The source category of the multi-source geological data corresponding to each causal evidence record and the line segment to which the bridge pile foundation's influence body belongs are calibrated; the same calibration calibrator is used for the same batch of bridge pile foundation influence bodies; the same threshold is used for the same multi-source geological data source category only when the data acquisition method, spatial accuracy level, overburden conditions, and the line segment to which it belongs are consistent or within the same calibration range. When the data exceeds the same calibration range, the corresponding causal evidence record is transferred to a restricted state or a pending verification state; if a new calibration calibrator has not yet been formed, and the necessary fields of the corresponding causal evidence record are complete, it is marked as a restricted state; if necessary fields are missing or cannot be converted, it is marked as a pending verification state.

[0137] Weights, overburden interference limiting thresholds, and effective applicable thresholds are bound using the same calibration version number for result traceability and version consistency control. During the identification of bridge pile foundation impact bodies in the current batch, already activated weights and thresholds remain frozen. After new borehole exposure data, construction process data, or supplementary geophysical data form new verification samples, they are only used in subsequent batches after recalibration and the generation of a new calibration version number. Output results for karst goaf identification do not automatically change with new calibration versions unless re-identification is performed on the corresponding bridge pile foundation impact bodies.

[0138] The specific generation process is as follows: When geophysical anomaly data is located below a thick overburden layer, or when its anomaly spatial range overlaps with pipelines, manhole covers, culverts, bridge abutment structures, road surface voids, loose backfill material, or metal reflection interference, the overburden interference is increased. If no contradiction is found between the geophysical anomaly and borehole data, mining data, or construction process data, the corresponding causal evidence record is marked as restricted, and the geophysical anomaly is added to the reserved objects. Causal evidence records under restricted status can still participate in the generation of causal candidate results, but the reason for restriction is recorded in their evidence source identifier, and the corresponding supplementary exploration priority is increased when supplementary exploration priority is generated.

[0139] When the borehole-revealed stratum is inconsistent with the geophysical anomaly depth stratum, and the depth deviation exceeds the allowable depth deviation, the evidence conflict trigger value of the corresponding causal evidence record is set to 1, and a conflict state is generated.

[0140] When the mining strata in the mining data correspond to deep anomalies, but the spatial overlap between the mining boundary projection and the bridge pile foundation influence body is lower than the lower limit of overlap specified according to the data spatial accuracy level, or the overlap between the remote sensing deformation range and the deep anomaly range is lower than the deformation response conflict threshold, the corresponding causal evidence record will also be marked as conflicting. Causal evidence records in conflicting states are not deleted, but are sent to evidence conflict processing to generate exploration conflict types, deformation response conflict types, composite causal conflict types, interference interpretation conflict types, or construction feedback conflict types.

[0141] When mining data lacks precision regarding mined-out boundaries, historical mining maps cannot be converted to coordinates, remote sensing deformation data lacks acquisition time, geophysical anomaly data lacks spatial precision identifiers, or borehole data cannot correspond to specific bridge pile foundation impact bodies, the verification gap trigger value of the corresponding causal evidence record is set to 1, and a pending verification status is generated. The pending verification status indicates that the evidence cannot yet be used as a sole basis for determining the causal category, but it can be used in conjunction with valid or limited evidence to indicate data coverage gaps.

[0142] A valid state is generated when there are no conflicting or unverified evidence records, the amount of overburden interference is below the overburden interference limitation threshold, and the comprehensive applicability value is not lower than the effective applicability threshold. Valid state causal evidence records can be directly used to match deep anomalies with karst causal evidence, mining subsidence causal evidence, or hydrological disturbance evidence. When multiple valid state evidence exist within the same bridge pile foundation influence body, they are compared sequentially according to depth-stratum correspondence, spatial overlap, temporal validity, and data accuracy reliability. Causal evidence records with higher values ​​are cited first. If the values ​​of the first item are the same, the values ​​of the next item are compared.

[0143] When new borehole data, construction process data, or supplementary geophysical data enter the engineering identification coordinate system, the data source applicability identifier is regenerated for the bridge pile foundation affected by the new data and for bridge pile foundations with spatial or stratigraphic correlations among their adjacent pile numbers. Updates use a single bridge pile foundation as the smallest window and the pile group under the same abutment as the linked window. When the same causal evidence record exhibits repeated states during continuous updates, a more conservative state is adopted as the current output, with conflicting states having higher priority than pending verification states, pending verification states having higher priority than restricted states, and restricted states having higher priority than valid states. In cases of coordinate registration failure, inconsistent depth benchmarks, missing construction record times, or inability to convert data accuracy levels, the data source is not marked as valid but is moved to pending verification states. If this abnormal condition simultaneously leads to a contrary conclusion with existing valid evidence, it is moved to a conflicting state.

[0144] Example 2:

[0145] Based on Example 1, this example illustrates the generation and recording method of evidence conflict types. This example uses deep anomalies, candidate causal results, and causal evidence records within the same bridge pile foundation influence body as conflict discrimination objects. Corresponding evidence conflict types are generated for inconsistencies in exploration and uncovering, inconsistencies in deformation response, overlapping of multiple causes, overlapping of shallow interferences, and inconsistencies in construction feedback. The candidate causal results involved in the conflict, the evidence source identifier, the conflict occurrence layer, and the bridge pile foundation influence body number are written into the evidence conflict record.

[0146] In this embodiment, the types of evidence conflict generated include:

[0147] When geophysical anomaly data and borehole exposure data are inconsistent in at least one of the object type, exposure status and depth stratum within the same bridge pile foundation influence body, an exploration-exposure conflict type is generated.

[0148] When the deformation range shown by remote sensing deformation data is inconsistent with the spatial range of deep anomalies, a deformation response conflict type is generated.

[0149] When mining data and karst genesis evidence are simultaneously associated with the same bridge pile foundation influence body, and both have a corresponding relationship with at least one of the influence ranges of the same deep anomaly and the same pile end bearing layer, a composite genetic conflict type is generated.

[0150] When shallow interference evidence overlaps with deep anomalies within a planar range, and the influence range of shallow interference evidence corresponds to the interpretability range of deep anomalies, an interference interpretation conflict type is generated.

[0151] When at least one of the abnormal layer, abnormal object and abnormal state shown in the construction process data is inconsistent with the existing candidate results of the cause, a construction feedback conflict type is generated.

[0152] In one implementation, after the evidence conflict type is generated, the evidence conflict record retains the candidate results of the causes of the conflict, the evidence source identifier, the source of the conflict data, the layer where the conflict occurred, and the number of the bridge pile foundation affected body.

[0153] When there is a conflict between exploration and exposure types, retain the anomaly response type corresponding to the geophysical anomaly data, the exposure object and exposure status corresponding to the borehole exposure data; when there is a conflict between deformation response types, retain the deformation range, deformation continuation status and deep anomaly range corresponding to the remote sensing deformation data.

[0154] When there are complex genetic conflict types, the evidence conflict record retains both karst genetic candidate results and goaf genetic candidate results, and records the evidence source identifiers corresponding to mining data, karst genetic evidence and geophysical anomaly data in the evidence conflict record;

[0155] When there are conflicting interpretation types, the evidence conflict record retains shallow interfering evidence, deep anomalies and corresponding interference categories, and writes deep anomalies as retained objects into the evidence conflict record;

[0156] When there is a construction feedback conflict type, the construction anomaly shown in the construction process data is converted into construction feedback evidence, and the construction feedback evidence is associated with the bridge pile foundation affected by the construction anomaly.

[0157] When a construction anomaly is spatially or hierarchically related to the impact body of the adjacent bridge pile foundation determined based on the adjacent pile relationship data, the construction feedback evidence will be synchronously associated with the impact body of the bridge pile foundation corresponding to the adjacent pile number.

[0158] In a preferred embodiment of Example 2, when generating evidence conflict types, deep anomalies, candidate causal results, and causal evidence records within the same bridge pile foundation influence body are used as conflict discrimination objects. A conflict discrimination object consists of a deep anomaly, a candidate causal result, and causal evidence records involved in generating the candidate causal result that are interconnected within the same bridge pile foundation influence body; when one deep anomaly corresponds to multiple candidate causal results, multiple conflict discrimination objects are formed respectively. Evidence conflict processing does not delete already generated candidate causal results, but instead writes contradictory data sources, contradictory directions, affected bridge pile foundation influence bodies, and subsequent processing actions into the evidence conflict record, thus forming a traceable data processing chain between karst genesis, mining subsidence genesis, hydrological disturbance, and shallow disturbance.

[0159] For exploration-discovery conflict types, an exploration-discovery conflict type is generated when geophysical anomaly data indicates the existence of deep anomalies within the influence body of a bridge pile foundation, but borehole data does not reveal karst caves, karst fracture zones, goaf boundaries, caving zones, or water-rich anomalies at the corresponding strata, or when the depth of the borehole revealed object deviates from the depth of the geophysical anomaly by more than the allowable error range. The degree of depth deviation is determined by the following formula:

[0160] ,

[0161] in, For the first The deviation of the exploration depth of each conflict-detecting object; For the first The center depth of the geophysical anomaly corresponding to each conflict discrimination object; For the first Each conflict detection object corresponds to the center depth of the borehole-revealed object; For the first The permissible depth error for exploration corresponding to each conflict discrimination object. The calibration is jointly determined by borehole depth measurement error, geophysical inversion vertical resolution, and formation undulation amplitude. The larger the value, the stronger the stratigraphic contradiction between geophysical anomalies and borehole revelation. When When the exploration conflict threshold is reached, or when the geophysical anomaly response type is opposite to the borehole exposure status, an exploration conflict type is triggered. After this type is triggered, both geophysical anomaly data and borehole exposure data are retained, along with the corresponding karst genesis candidate results, mining subsidence gene candidate results, or hydrological disturbance candidate results. The output fields include at least the bridge pile foundation influence body number, evidence source identifier of the geophysical anomaly data, borehole number, anomaly response type, exposure object, exposure status, exploration depth deviation, and exploration conflict type.

[0162] For deformation response conflict types, a deformation response conflict type is generated when the deformation range given by the remote sensing deformation data does not match the planar projection range of the deep anomaly, or when the deep anomaly has formed a candidate causal result, and the remote sensing deformation data has a calculable boundary and deformation continuation state, but does not show a continuous deformation response consistent with the interpretation result of the deep anomaly. If the remote sensing deformation data does not have a calculable boundary or deformation continuation state, the corresponding causal evidence record is marked as pending verification, and a deformation response conflict type is not triggered independently.

[0163] The correspondence between the deformation range and the deep anomaly is determined by the following formula:

[0164] ,

[0165] in, For the first The amount of deformation anomaly overlap of each conflict-detecting object; This is a function for calculating the area of ​​a planar region. For the first The planar projection area of ​​the remote sensing deformation range corresponding to each conflict discrimination object; For the first Each conflict-detection object corresponds to a planar projection region of a deep anomaly. The value ranges from 0 to 1; the smaller the value, the weaker the spatial response relationship between remote sensing deformation and deep anomalies. When or If there is a lack of calculable boundaries, or if the corresponding denominator area is zero, the amount of deformation anomaly overlap is not calculated, and the causal evidence record corresponding to the conflict discrimination object is marked as pending verification.

[0166] The deformation response conflict threshold is calibrated based on remote sensing deformation resolution, the planar dimensions of the bridge pile foundation influence body, and the historical settlement influence radius; when When the deformation response conflict threshold is lower than the threshold and the remote sensing deformation persistence state is inconsistent with the interpretation results of deep anomalies, a deformation response conflict type is triggered. After this type is triggered, the remote sensing deformation data, deep anomalies, and generated candidate causal results are retained. The output fields include at least the bridge pile foundation influence body number, the evidence source identifier of the remote sensing deformation data, the deformation range, the deformation persistence state, the range of deep anomalies, the amount of deformation anomaly overlap, the direction of response inconsistency, and the deformation response conflict type. The direction of response inconsistency includes three types: deformation without corresponding deep anomalies, deep anomalies without deformation response, and deformation range offset.

[0167] For composite genetic conflict types, the degree of composite superposition is determined by the proportion of the intersection volume of karst-originating candidate results and mining-originating candidate results within the influence body of the bridge pile foundation to the combined volume of the two within that influence body. The degree of composite superposition is determined by the following formula:

[0168] ,

[0169] in, For the first The number of overlapping composite causes of conflict-discriminating objects; This is a function for calculating spatial volume. For the first The spatial range of karst genesis candidate results corresponding to each conflict discrimination object; For the first The spatial range of the candidate results for the cause of mining for each conflict discrimination object; For the first The spatial extent of the bridge pile foundation influence body to which the conflict determination object belongs. If the corresponding volume in the denominator is zero, or if any spatial extent is missing from the karst origin candidate result, mining subsidence candidate result, or bridge pile foundation influence body, then it is not determined based on... Trigger a compound cause conflict type and mark the corresponding cause evidence record as pending verification.

[0170] The value ranges from 0 to 1. A larger value indicates a higher degree of overlap between karst genesis candidate results and mining subsidence gene candidate results within the same engineering identification unit. The composite genetic conflict threshold is calibrated based on overlapping samples from confirmed karst development areas, historical mining areas, and pile end bearing strata. A composite genetic conflict type is triggered when the composite genetic conflict threshold is reached, or when mining data and karst genetic evidence both point to the same deep anomaly. Once triggered, this type does not involve a single selection between karst genetic candidate results and mining subsidence genetic candidate results; instead, both are retained simultaneously, and the sources of mining data, borehole exposure data, geophysical anomaly data, hydrogeological data, and engineering geological survey data are recorded in the evidence source identifier. Output fields must include at least the bridge pile foundation impact body number, karst genetic candidate result number, mining subsidence genetic candidate result number, evidence source identifier for mining data, evidence source identifier for borehole exposure data, evidence source identifier for geophysical anomaly data, composite genetic overlay amount, and composite genetic conflict type.

[0171] For interference interpretation conflict types, an interference interpretation conflict type is generated when the planar projection of shallow interference evidence overlaps with the planar projection of deep anomalies, the anomaly response of the corresponding geophysical method is susceptible to the influence of the shallow interference evidence, and there is a stratigraphic correspondence between the influence depth range of the shallow interference evidence and the interpretation depth of the deep anomaly. If the shallow interference evidence only overlaps in the plane, but its influence depth range is separate from the interpretation depth of the deep anomaly, or the corresponding geophysical method can distinguish between shallow and deep responses, then an interference interpretation conflict type is not triggered, and only the reason for the limitation of shallow interference is recorded in the evidence source identifier.

[0172] The projected coverage of shallow disturbances and deep anomalies is determined by the following formula:

[0173] ,

[0174] in, For the first The interference coverage of each conflict-detection object; This is a function for calculating the area of ​​a planar region. For the first Each conflict-detection object corresponds to a planar projection area of ​​shallow interference evidence; For the first Each conflict-detection object corresponds to a planar projection region of a deep anomaly. The value ranges from 0 to 1; the larger the value, the stronger the influence of shallow disturbances on the interpretation of deep anomalies. or If a calculable boundary is missing, or the corresponding denominator area is zero, the interference coverage is not calculated, and the causal evidence record corresponding to the conflict discrimination object is marked as pending verification. If only the shallow interference evidence lacks a boundary while the deep anomaly boundary is complete, the deep anomaly and causal candidate results are retained, and the interference interpretation conflict type is not triggered by the shallow interference evidence.

[0175] The interference coverage threshold is determined based on the detection depth of the geophysical method, the distribution density of shallow structures, and the calibration of verified interference samples; when When the interference coverage threshold is reached and the data source applicability identifier of the shallow interference evidence is in a valid or restricted state, an interference interpretation conflict type is triggered. After this type is triggered, both deep anomalies and shallow interference evidence are retained, with the deep anomaly written as a retained object into the evidence conflict record. Output fields include at least the bridge pile foundation impact body number, shallow interference category, evidence source identifier of the shallow interference evidence, evidence source identifier corresponding to the deep anomaly, interference coverage, anomaly response type, retained interpretation direction, and interference interpretation conflict type. The retained interpretation direction includes three types: priority verification of deep anomalies, priority verification of shallow interference, and simultaneous verification of both sources.

[0176] For construction feedback conflict types, a construction feedback conflict type is generated when the drilling status, grout leakage record, drill drop record, hole collapse record, water inrush record, mud level change record, drill cuttings exposure record, or actual rock entry status record in the construction process data are inconsistent with the existing candidate results of the current bridge pile foundation impact body.

[0177] After a construction feedback conflict is triggered, the scope of the updated objects is determined by the following formula:

[0178] ,

[0179] in, For the first A set of construction feedback update objects corresponding to each conflict detection object; For the first The bridge pile foundation influence body to which the conflict determination object belongs; In order to be with the first The first conflict-determining object has an adjacent stake relationship. One bridge pile foundation affected body; For the first The pile number corresponding to the affected body of each bridge pile foundation; For the first The set of adjacent pile numbers recorded in the bridge pile foundation impact body record to which each conflict determination object belongs. The bridge pile foundation impact body corresponding to the adjacent pile number is only related to the pile tip bearing layer, construction anomaly depth, or deep anomaly layer. When a conflict identification object has a corresponding relationship, it participates in the identification result update; when the above corresponding relationship is not met, the adjacent association is only recorded in the evidence source identifier, and the original cause category of the adjacent bridge pile foundation influence body is kept in the reserved state.

[0180] When construction feedback data is consistent with existing causal candidate results, an evidence source identifier corresponding to the construction process data is added to the evidence source identifier, and the causal category corresponding to the causal candidate result is maintained. When construction feedback data contradicts existing causal candidate results, the original causal candidate result is retained and a construction feedback conflict type is generated. When the construction process data lacks at least one of time, station number, and depth stratum, the corresponding construction feedback evidence record is marked as pending verification, and the original causal candidate result is kept in a retained state. The output fields of this type include at least the construction station number, construction anomaly type, construction anomaly depth, evidence source identifier of construction process data, original causal candidate result, updated identification result, affected adjacent station number, supplementary investigation priority adjustment direction, and construction feedback conflict type.

[0181] To avoid duplicate interpretations of conflicts between different data sources within the same bridge pile foundation impact body, conflict identification objects are established based on the correspondence between deep anomalies, candidate causal results, and causal evidence records. When a deep anomaly corresponds to multiple candidate causal results, multiple conflict identification objects are generated separately; when a candidate causal result corresponds to multiple evidence source identifiers, all evidence source identifiers are retained, and the data source applicability identifier corresponding to each evidence source identifier is recorded. Evidence conflict records do not change the original data content; only the conflict type, triggering reason, retained object, and subsequent review direction are recorded. The output records for the five types of evidence conflict use a unified field structure, including conflict record number, bridge pile foundation impact body number, conflict type, triggering condition, triggering evidence source identifier, retained object, conflict quantification value, evidence source identifier, candidate causal result number, risk level adjustment direction, supplementary investigation priority adjustment direction, and matters to be reviewed. Risk level adjustment directions include upward, maintain, and downward adjustment; supplementary investigation priority adjustment directions include upward, maintain, and decrease. When multiple conflict types are triggered simultaneously, processing actions are written in the following order: construction feedback conflict type, exploration conflict type, composite causal conflict type, interference interpretation conflict type, and deformation response conflict type. The conflict quantification value refers to the numerical field written to the evidence conflict record along with the evidence conflict type. The conflict quantification value corresponding to the exploration conflict type is the exploration depth deviation; the conflict quantification value corresponding to the deformation response conflict type is the deformation anomaly overlap; the conflict quantification value corresponding to the composite causal conflict type is the composite causal overlay; the conflict quantification value corresponding to the interference interpretation conflict type is the interference coverage; and the conflict quantification value corresponding to the construction feedback conflict type is the deviation between the construction anomaly depth and the depth layer corresponding to the existing causal candidate result. When a calculable boundary, depth benchmark, or necessary source field is missing, the evidence conflict record is written with the item to be verified, and the conflict quantification value field is set to null.

[0182] When multiple conflict types trigger simultaneously and point to the same output field, the outputs are merged according to the conservative direction. When there are multiple adjustment directions for risk level (upward, maintain, and downward), upward adjustment is adopted. When there are multiple adjustment directions for supplementary exploration priority (increase, maintain, and decrease), increase is adopted. When the conflict state or pending verification state has not been resolved, the processing method of lowering the risk level or reducing the supplementary exploration priority is not adopted, and the risk level or supplementary exploration priority is not lowered because a single data source does not show anomalies.

[0183] Candidate results within the same bridge pile foundation influence area are retained according to their original causal category, and corresponding conflict type and evidence source identifiers are added to the karst goaf identification results. The above order is only used to determine the sequence of review and update, and is not used to delete any triggered causal candidate results.

[0184] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

[0185] Furthermore, those skilled in the art will understand that although some embodiments herein include certain features included in other embodiments but not others, combinations of features from different embodiments are meant to be within the scope of this application and form different embodiments. For example, all the embodiments above can be used in any combination. The information disclosed in this background section is intended only to enhance the understanding of the general background of this application and should not be construed as an admission or in any way implying that such information constitutes prior art known to those skilled in the art.

Claims

1. A method for identifying karst goaf areas based on multi-source data fusion, characterized in that, include: The bridge pile foundation influence body is generated based on bridge pier data, abutment data, pile foundation design data, pile end bearing layer data, and adjacent pile relationship data. The bridge pile foundation influence body is a three-dimensional engineering identification unit covering the pile body, pile end bearing layer, and the associated range of adjacent piles. Acquire multi-source geological data corresponding to the bridge pile foundation influence body and register it to an engineering identification coordinate system indexed by route mileage, lateral offset, and stratum depth; Genetic evidence records, including karst genesis evidence, mining subsidence evidence, hydrological disturbance evidence, and shallow disturbance evidence, are generated within each bridge pile foundation impact area. Data source applicability identifiers are configured to distinguish between valid, restricted, conflicting, and pending verification states based on the overburden disturbance state, depth-layer correspondence, spatial overlap, acquisition time, and data accuracy level. Based on shallow interference evidence, geophysical anomaly data are marked for interference. The retained deep anomalies are matched with karst genesis evidence, mining subsidence evidence, and hydrological disturbance evidence to generate genetic candidate results. When multiple candidate causal results correspond to the same bridge pile foundation affected body and the causal categories are mutually exclusive, an evidence conflict type is generated and the corresponding candidate causal results are retained. Output the karst goaf identification results of the bridge pile foundation affected body. The karst goaf identification results include the causal category, risk level, evidence source identifier, evidence conflict type, and supplementary investigation priority.

2. The method for identifying karst goaf based on multi-source data fusion according to claim 1, characterized in that, The multi-source geological data includes geophysical anomaly data, as well as at least two of the following: engineering geological survey data, borehole exposure data, mining data, remote sensing deformation data, hydrogeological data, and construction process data. When each type of multi-source geological data enters the engineering identification coordinate system, it generates an evidence source identifier, a collection time identifier, and a spatial accuracy identifier.

3. The method for identifying karst goaf based on multi-source data fusion according to claim 1, characterized in that, The generation of the bridge pile foundation influence body includes: A single pile influence body is generated based on a single pile, a pile group influence body is generated based on multiple piles under the same pier cap, and a pier influence body is generated based on the planar projection relationship of adjacent pier caps and the continuous range of the bearing stratum at the pile tip. Each bridge pile foundation influence body records the pier number, pile number, pile tip elevation, bearing stratum range, and adjacent pile number.

4. The method for identifying karst goaf based on multi-source data fusion according to claim 1, characterized in that, Registering the multi-source geological data to the engineering identification coordinate system includes: The data obtained from borehole exposure are converted into exposure strata, exposure objects, and exposure status; geophysical anomaly data are converted into anomaly spatial range, anomaly depth strata, and anomaly response type; mining data are converted into mining range, mining strata, and mining boundaries; remote sensing deformation data are converted into deformation range and deformation continuity status; and hydrogeological data are converted into groundwater activity strata and water-rich status.

5. The method for identifying karst goaf based on multi-source data fusion according to claim 1, characterized in that, The evidence for karst formation includes at least one of the following: distribution of soluble rocks, karst landforms, exposure of caves, exposure of dissolution fracture zones, and groundwater connectivity. The evidence for the formation of the goaf includes at least one of the following: mining area, mining strata, goaf boundary, rift zone, and historical subsidence; The evidence of hydrological disturbance includes at least one of groundwater level changes, water abundance anomalies, water inrush records, and grout leakage records. The shallow interference evidence includes at least one of pipelines, manhole covers, culverts, bridge abutment structures, road surface voids, loose backfill material, and metal reflection interference.

6. The method for identifying karst goaf based on multi-source data fusion according to claim 1, characterized in that, The generated causal candidate results include: When a deep anomaly corresponds to at least one of the following: distribution of soluble rocks, karst landforms, exposure of caves, exposure of dissolution fracture zones, and groundwater connectivity, a karst genesis candidate result is generated. When a deep anomaly corresponds to at least one of the following: mining area, mining stratum, goaf boundary, caving zone, and historical subsidence, a goaf genesis candidate result is generated. When a deep anomaly corresponds to at least one of groundwater level changes, water-rich anomalies, water inrush records, and grout leakage records, a candidate hydrological disturbance result is generated.

7. The method for identifying karst goaf areas based on multi-source data fusion according to claim 1, characterized in that, Types of evidence conflict include: When geophysical anomaly data and borehole exposure data are inconsistent in at least one of the object type, exposure status and depth stratum within the same bridge pile foundation influence body, an exploration-exposure conflict type is generated. When the deformation range shown by remote sensing deformation data is inconsistent with the spatial range of deep anomalies, a deformation response conflict type is generated. When mining data and karst genesis evidence are simultaneously associated with the same bridge pile foundation influence body, and both have a corresponding relationship with at least one of the influence ranges of the same deep anomaly and the same pile end bearing layer, a composite genetic conflict type is generated. When shallow interference evidence overlaps with deep anomalies within a planar range, and the influence range of shallow interference evidence corresponds to the interpretability range of deep anomalies, an interference interpretation conflict type is generated. When at least one of the abnormal layers, abnormal objects, and abnormal states shown in the construction process data is inconsistent with the existing candidate results for causes, a construction feedback conflict type is generated.

8. The method for identifying karst goaf based on multi-source data fusion according to claim 7, characterized in that, After generating the evidence conflict type, an evidence conflict record is generated. The evidence conflict record retains the candidate results of the cause of the conflict, the evidence source identifier, the triggering evidence source identifier, the conflict occurrence layer, and the number of the bridge pile foundation affected body. When there is a conflict type in the exploration and exposure, the evidence conflict record retains the abnormal response type corresponding to the geophysical exploration anomaly data, the exposure object corresponding to the borehole exposure data, and the exposure status. When a deformation response conflict type exists, the evidence conflict record retains the deformation range, deformation continuation status, and deep anomaly range corresponding to the remote sensing deformation data; When there is a complex genetic conflict type, the evidence conflict record simultaneously retains the karst genetic candidate results and the goaf genetic candidate results, and records the evidence source identifiers corresponding to the mining data, karst genetic evidence and geophysical anomaly data in the evidence conflict record; When there are conflicting interpretation types, the evidence conflict record retains shallow interfering evidence, deep anomalies and corresponding interference categories, and writes the deep anomalies as retained objects into the evidence conflict record; When there is a construction feedback conflict type, the construction anomaly shown in the construction process data is converted into construction feedback evidence, and the construction feedback evidence is associated with the bridge pile foundation affected by the construction anomaly. When a construction anomaly is spatially or hierarchically related to the impact body of adjacent bridge pile foundations determined based on adjacent pile relationship data, the construction feedback evidence will be synchronously associated with the impact body of the adjacent bridge pile foundations.

9. The method for identifying karst goaf based on multi-source data fusion according to claim 1, characterized in that, The risk levels include high risk, medium risk, and low risk; Candidate causal results intersecting with the pile tip bearing stratum are identified as high-risk; candidate causal results located within the lateral influence range of the pile body and intersecting with the associated range of adjacent piles are identified as medium-risk; and candidate causal results located at the edge of the bridge pile foundation influence body and separated from the pile tip bearing stratum are identified as low-risk. Data coverage gaps are gaps in multi-source geological data within the influence body of bridge pile foundations, resulting from at least one of the following: missing spatial range, missing depth and stratum, missing acquisition time, missing spatial accuracy identification, missing evidence source identification, and missing coordinate registration results. Supplementary exploration priorities are generated according to risk level, type of evidence conflict, and data coverage gaps.

10. The method for identifying karst goaf based on multi-source data fusion according to claim 2, characterized in that, The construction process data includes at least one of the following: drilling status data, grout leakage records, drill bit loss records, borehole collapse records, water inrush records, mud level change records, drill cuttings exposure records, and actual rock entry status records. When construction process data indicates that there is a construction anomaly in a certain bridge pile foundation affected body, the construction anomaly is converted into construction feedback evidence, and the karst mining area identification results, evidence conflict type, and supplementary investigation priority of the bridge pile foundation affected body and the adjacent bridge pile foundation affected bodies determined based on adjacent pile relationship data are updated.

Images