Method and system for analyzing occurrence of sparse underground ore body based on multi-dimensional sensing data

Abstract

The application relates to the technical field of rare earth ore body, and in particular to a rare earth underground ore body occurrence analysis method and system based on multidimensional sensing data. The method comprises the following steps: acquiring a multi-source heterogeneous sensing data set, eliminating the deviations of different sensing data in sampling scales, timestamps and spatial positions based on the multi-source heterogeneous sensing data set, and obtaining spatio-temporal alignment multi-source sensing data; acquiring a geological environment information set, combining the spatio-temporal alignment multi-source sensing data, jointly inverting and analyzing the extension trend of ion adsorption type rare earth ore body based on the geological environment information set, and obtaining ore body extension trend inversion information; based on the ore body extension trend inversion information, adaptively analyzing an abnormal section of the underground ore body occurrence, generating and outputting a rare earth underground ore body occurrence analysis log containing a three-dimensional result of the ore body occurrence. The application promotes the intelligent and accurate development of rare earth exploration and development, and helps the efficient and reasonable development of rare earth resources.

Description

Technical Field

[0001] This application relates to the field of rare earth ore body technology, and in particular to a method and system for analyzing the occurrence of rare earth underground ore bodies based on multidimensional sensor data. Background Technology

[0002] Currently, the detection and occurrence analysis of underground ore bodies, especially ion adsorption type rare earth ore bodies, mainly rely on single or limited geophysical and geological methods. Common practices include using independently acquired sensor data such as microseismic monitoring and electromagnetic exploration, combined with regional geological survey data for manual interpretation and inference. However, these methods have significant limitations in integrating multi-source data: different sensors often differ in time reference, spatial sampling scale, and observation location, making it difficult to directly fuse the data.

[0003] Existing methods suffer from significant biases in the identification and delineation of underground rare earth ore bodies due to fragmented multi-source data and disconnect from geophysical information. The spatiotemporal mismatch at the data level prevents the full utilization of the advantages of multi-sensor collaboration. Furthermore, the isolated state of information at the information level makes the final ore body delineation results uncertain and difficult to accurately reflect the three-dimensional spatial distribution of ore bodies, especially under complex occurrence conditions. This directly restricts the accuracy and reliability of exploration, which may not only lead to misjudgment of resources and risks in mining design, but also become a technical bottleneck for the efficient exploration of deep and concealed rare earth ore bodies. Summary of the Invention

[0004] This application provides a method and system for analyzing the occurrence of rare earth underground ore bodies based on multidimensional sensor data, in order to solve the above-mentioned problems.

[0005] Firstly, this application provides a method for rare earth underground orebody occurrence analysis based on multidimensional sensor data. The method includes: acquiring a multi-source heterogeneous sensor dataset; based on the multi-source heterogeneous sensor dataset, eliminating the biases of different sensor data in sampling scale, timestamp, and spatial location to obtain spatiotemporally aligned multi-source sensor data; acquiring a geological environment information set; based on the geological environment information set and combined with the spatiotemporally aligned multi-source sensor data, performing joint inversion analysis on the extension trend of ion-adsorption type rare earth orebody to obtain orebody extension trend inversion information; and based on the orebody extension trend inversion information, performing adaptive analysis on anomalous sections of underground orebody occurrence, generating and outputting a rare earth underground orebody occurrence analysis log containing three-dimensional orebody occurrence results.

[0006] The above technical solutions solve the core problem of spatiotemporal heterogeneity of multi-source sensor data in rare earth exploration, laying a standardized data foundation for ore body analysis. The joint inversion with geological environmental information makes the analysis of ore body extension trends more accurate and comprehensive. Adaptive analysis enables intelligent identification of abnormal sections. The three-dimensional results make the occurrence of ore bodies more intuitive. The output analysis log provides a scientific basis for rare earth exploration and mining decisions, improves analysis efficiency and accuracy, promotes the intelligent and precise development of rare earth exploration and development, and helps to develop rare earth resources efficiently and rationally.

[0007] Optionally, the step of eliminating deviations in sampling scale, timestamps, and spatial locations of different sensor data based on the multi-source heterogeneous sensor dataset to obtain spatiotemporally aligned multi-source sensor data includes: the multi-source heterogeneous sensor dataset includes microseismic response data and electromagnetic detection data; based on the microseismic response data, discrete event timestamps are analyzed, and synchronization is performed by introducing a unified time reference event to obtain time-unified microseismic data; based on the electromagnetic detection data, the coordinates of multi-source measuring points are analyzed, and scale normalization is performed through a unified grid spatial resampling rule to obtain spatial grid electromagnetic data; based on the time-unified microseismic data and combined with the spatial grid electromagnetic data, the spatial coordinates and temporal attributes of the two are registered and fused to obtain the spatiotemporally aligned multi-source sensor data.

[0008] Optionally, the step of registering and fusing the spatial coordinates and temporal attributes of the time-unified microseismic data and the spatial grid electromagnetic data to obtain the spatiotemporally aligned multi-source sensing data includes: analyzing the spatial coordinate distribution of microseismic events based on the time-unified microseismic data; mapping the coordinates of each microseismic event to the corresponding spatial grid cell based on the unified grid space followed by the spatial grid electromagnetic data to obtain a spatially registered microseismic event sequence; analyzing the unified time reference corresponding to each event based on the microseismic event sequence; performing time window association matching between the microseismic events and the electromagnetic detection results based on the detection time interval of the spatial grid electromagnetic data on the corresponding grid cell to obtain a set of spatiotemporally associated event pairs; and analyzing the correspondence between the microseismic event attributes and electromagnetic response characteristics within the same grid cell based on the set of spatiotemporally associated event pairs. By superimposing and verifying the consistency of the two features, a feature description of the fused microseismic and electromagnetic information within each grid cell is generated to obtain the spatiotemporally aligned multi-source sensing data.

[0009] Optionally, the joint inversion analysis of the extension trend of ion-adsorption type rare earth ore bodies based on the geological environment information set and combined with the spatiotemporally aligned multi-source sensing data to obtain ore body extension trend inversion information includes: the geological environment information set includes hydrogeological information and ore body weathering crust information; based on the hydrogeological information, the constraint of groundwater runoff on the ion migration and enrichment law of the ore body is analyzed to obtain groundwater constraint information; based on the ore body weathering crust information, the constraint of weathering crust stratification on the ore body basement morphology and three-dimensional distribution framework is analyzed to obtain weathering crust morphology information; and the groundwater constraint information is integrated with the... Based on the weathering crust morphology information, a basic geological spatial constraint framework for the ore body is constructed. Using the spatiotemporally aligned multi-source sensor data, the interaction between microseismic event attributes and electromagnetic response characteristics is analyzed. According to the synergistic variation information of spatial attenuation of microseismic energy and anomalous response of electromagnetic apparent resistivity in the ore body occurrence area, a multi-source synergistic anomaly feature set is obtained. Based on the basic geological spatial constraint framework, combined with the multi-source synergistic anomaly feature set, the coupling relationship between geological constraints and geophysical anomalies in three-dimensional space is analyzed. The extension boundary and trend of the ore body are optimized, deduced, and delineated, resulting in the inversion information of the ore body's extension trend.

[0010] Optionally, the construction process of the multi-source coordinated anomaly feature set includes: based on the time-unified microseismic data, analyzing the energy attenuation gradient characteristics of microseismic event attributes in the spatial propagation process within each spatial grid cell to obtain microseismic energy attenuation information; based on the spatial grid electromagnetic data, analyzing the apparent resistivity change characteristics of electromagnetic response features within the same spatial grid cell in three-dimensional space to obtain electromagnetic apparent resistivity anomaly information; based on the microseismic energy attenuation information and combined with the electromagnetic apparent resistivity anomaly information, analyzing the synchronicity of their changes within the same spatial grid cell, and identifying coordinated anomaly zones related to ore body occurrence based on the regions where the increase in microseismic energy attenuation gradient and the decrease in electromagnetic apparent resistivity anomaly occur simultaneously in space; based on the coordinated anomaly zones, analyzing the corresponding intensity and spatial continuity of the microseismic energy attenuation characteristics and the electromagnetic apparent resistivity anomaly characteristics, and obtaining the multi-source coordinated anomaly feature set through joint measurement and spatial aggregation of the two characteristics.

[0011] Optionally, the specific implementation of analyzing the coupling relationship between geological constraints and geophysical anomalies in three-dimensional space includes: based on the microseismic energy attenuation information, analyzing the gradient change characteristics along the depth direction within each spatial grid cell to obtain microseismic gradient change information; based on the electromagnetic apparent resistivity anomaly information, analyzing the anomaly fluctuation characteristics with increasing depth within the same spatial grid cell to obtain electromagnetic anomaly fluctuation information; based on the microseismic gradient change information and combined with the electromagnetic anomaly fluctuation information, analyzing the matching degree of the changing trends of the two in the corresponding spatial grid cell and depth interval, and identifying synchronous change areas based on the degree of synergy between gradient increase and anomaly decrease in spatial location and change pattern; based on the synchronous change areas, analyzing the continuity and correlation strength of microseismic gradient and electromagnetic anomaly fluctuations in three-dimensional space to obtain the coupling relationship.

[0012] Optionally, the optimization and delineation of the ore body's extension boundary and trend to obtain the ore body extension trend inversion information includes: based on the synchronous change region, analyzing the spatial stability and consistency of the microseismic gradient change information and the electromagnetic anomaly fluctuation information to obtain mineralization clue regions; according to the coupling relationship, analyzing the continuity and change trend of the mineralization clue regions in three-dimensional space, and combining the basic geological spatial constraint framework to infer the main extension direction of the ore body in the horizontal and vertical directions to obtain the ore body three-dimensional morphology deduction information; based on the ore body three-dimensional morphology deduction information and the multi-source collaborative anomaly feature set, within the boundary of the basic geological spatial constraint framework, smoothly connecting and optimizing the spatial contour of the ore body to delineate the continuous distribution range of the ore body in three-dimensional space to obtain the ore body extension trend inversion information.

[0013] Optionally, the step of adaptively analyzing anomalous sections of underground orebody occurrence based on the orebody extension trend inversion information to generate and output a rare earth underground orebody occurrence analysis log containing three-dimensional orebody occurrence results includes: analyzing local abrupt changes and discontinuities in morphology, thickness, and geophysical response characteristics within the three-dimensional distribution range of the orebody based on the orebody extension trend inversion information to obtain orebody occurrence anomalous section information; obtaining orebody occurrence information based on the orebody occurrence anomalous section information, combined with the basic geological spatial constraint framework, and according to adaptive weighted correction of the anomalous causes; and integrating the orebody extension trend inversion information, the orebody occurrence anomalous section information, and the corresponding orebody occurrence information to generate and output the rare earth underground orebody occurrence analysis log.

[0014] Optionally, the process of constructing the ore body occurrence information includes: based on the ore body occurrence anomaly segment information, analyzing the characteristics of each anomaly segment in terms of morphological abrupt change, thickness change rate, and geophysical response instability to obtain occurrence anomaly characteristic information; based on the occurrence anomaly characteristic information, combined with the hydrogeological information and the ore body weathering crust information, analyzing the correlation between different anomaly characteristics and geological environmental genesis to obtain anomaly gene discrimination information; based on the anomaly gene discrimination information, combined with the ore body three-dimensional morphology inference information, and according to the influence weights of different genesis factors such as hydrological activity, weathering crust interface changes, and tectonic interference on the continuity of ore body occurrence, differentially weighted correction is performed on the occurrence inference results of the anomaly segments to obtain corrected ore body occurrence information; based on the corrected ore body occurrence information, combined with the basic geological spatial constraint framework, analyzing the coordination between the correction results and the overall geological framework, and obtaining the final coordinated ore body occurrence information through feedback verification with the multi-source collaborative anomaly feature set.

[0015] Secondly, this application provides a rare earth underground ore body occurrence analysis system based on multidimensional sensor data. The system includes: a data alignment module for acquiring a multi-source heterogeneous sensor dataset, and based on the multi-source heterogeneous sensor dataset, eliminating the deviations of different sensor data in sampling scale, timestamp, and spatial location to obtain spatiotemporally aligned multi-source sensor data; an inversion analysis module for acquiring a geological environment information set, and based on the geological environment information set, combined with the spatiotemporally aligned multi-source sensor data, performing joint inversion analysis on the extension trend of ion-adsorption type rare earth ore bodies to obtain ore body extension trend inversion information; and an adaptive analysis module for performing adaptive analysis on abnormal sections of underground ore body occurrence based on the ore body extension trend inversion information, generating and outputting a rare earth underground ore body occurrence analysis log containing three-dimensional ore body occurrence results. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a schematic diagram illustrating an application scenario provided in one embodiment of this application;

[0018] Figure 2 A flowchart of a rare earth underground ore body occurrence analysis method based on multidimensional sensor data provided in an embodiment of this application;

[0019] Figure 3This is a schematic diagram of the structure of a rare earth underground ore body occurrence analysis system based on multidimensional sensor data, provided in an embodiment of this application. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0021] Furthermore, the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. Additionally, the character " / " in this article, unless otherwise specified, generally indicates that the preceding and following related objects have an "or" relationship.

[0022] The embodiments of this application will now be described in further detail with reference to the accompanying drawings.

[0023] In the analysis of occurrence of ion-adsorption type rare earth underground ore bodies based on multidimensional sensor data, the current ion-adsorption type rare earth exploration methods are limited and singular. The spatiotemporal benchmarks and sampling scales of various sensor data are different, making it difficult to integrate and communicate. The data is fragmented and the geological and physical information is disconnected, resulting in significant deviations in ore body identification and insufficient exploration accuracy, which seriously restricts the development of deep concealed ore body exploration technology.

[0024] Based on this, this application provides a method and system for analyzing the occurrence of rare earth underground ore bodies based on multi-dimensional sensor data, breaking down barriers to multi-source data fusion and unifying rare earth exploration data standards; it also links geological data to deepen ore body analysis and intelligently screen risk zones. The system presents a three-dimensional view of the ore body morphology, leaves data traces throughout the entire process to support mining decisions, significantly reduces costs and increases efficiency, and facilitates the refined development and layout of rare earth resources.

[0025] Figure 1 This application provides an illustration of an application scenario. In the process of analyzing the occurrence of ion-adsorption type rare earth underground ore bodies based on multidimensional sensor data, the method provided in this application is used to build a standardized data base; geological information is integrated for joint inversion, anomalies are intelligently identified using adaptive algorithms, and the overall ore body is displayed in three-dimensional visualization, which helps to promote intelligent upgrades in exploration and ensures the scientific and efficient development and utilization of rare earth resources.

[0026] Specifically, the method provided in this application can be applied to any server, where the server interacts with ore body detection sensing equipment and historical drilling logs to acquire multi-source heterogeneous sensor datasets provided by the ore body detection sensing equipment and geological environment information sets provided by historical drilling logs. This addresses the technical challenge of heterogeneous spatiotemporal data in rare earth exploration, optimizes ore body projection accuracy through joint geological inversion, and enables intelligent anomaly identification using adaptive algorithms. Three-dimensional modeling visually reconstructs the current state of the ore body, generating and outputting rare earth underground ore body occurrence analysis logs for rare earth ore body detection personnel. This end-to-end data traceability strengthens the foundation for decision-making and empowers the industry's intelligent transformation and scientific resource development.

[0027] For specific implementation details, please refer to the following examples.

[0028] Figure 2 This is a flowchart illustrating a method for analyzing the occurrence of rare earth underground ore bodies based on multidimensional sensor data, provided in one embodiment of this application. The method of this embodiment can be applied to servers in the above scenarios. Figure 2 As shown, the method includes:

[0029] S201. Obtain multi-source heterogeneous sensor dataset. Based on the multi-source heterogeneous sensor dataset, eliminate the biases of different sensor data in sampling scale, timestamp and spatial location to obtain spatiotemporally aligned multi-source sensor data.

[0030] Multi-source heterogeneous sensor datasets can be collections of rare earth orebody detection-related data of different types and specifications, differing in data structure, acquisition attributes, and expression forms, generated during the exploration of ion-adsorption type rare earth mining areas. The orebody detection sensor equipment serves as the data source. The sampling scale can be the spatial interval standard for the sensor equipment to collect data on the underground geological environment and orebody information of the rare earth mining area. The timestamp can be the time stamp information recorded by the sensor equipment when collecting orebody-related data. The spatial location can be the underground geographic coordinates corresponding to the data acquisition by the sensor equipment. Spatiotemporally aligned multi-source sensor data can be multi-source sensor data that has been processed from the original multi-source heterogeneous sensor dataset to eliminate deviations in sampling scale, timestamp, and spatial location, resulting in multi-source sensor data with unified spatial acquisition standards, unified time recording benchmarks, and a unified spatial coordinate system.

[0031] Specifically, in the exploration of underground ore bodies of ion-adsorption type rare earth deposits, the differences in specifications and deployment of multiple sensing devices result in deviations in sampling scale, timestamps, and spatial positions, making it impossible to directly fuse data. This can easily lead to distorted analysis results and restrict accurate exploration. By acquiring raw data through data acquisition terminals and using calibration and registration methods, unified processing of sampling, time, and space can be completed to obtain standardized spatiotemporally aligned multi-source sensor data, providing data support for subsequent analysis.

[0032] S202. Obtain the geological environment information set. Based on the geological environment information set and combined with spatiotemporally aligned multi-source sensor data, conduct joint inversion analysis on the extension trend of ion adsorption type rare earth ore bodies to obtain ore body extension trend inversion information.

[0033] The geological environment information set can be a collection of various information comprehensively reflecting the geological background characteristics of an ion-adsorption type rare earth mining area, compiled from historical drilling data. Ion-adsorption type rare earth ore bodies can be rare earth mineral bodies where rare earth elements exist in the weathering crust of the mining area in the form of ion adsorption. The extension trend can be the strike, dip, and dip angle changes of ion-adsorption type rare earth ore bodies in underground space, as well as the spatial distribution characteristics such as the extension range and thickness variations of the ore body in three-dimensional space. Joint inversion analysis can be a comprehensive analytical method that combines multiple data sources and analytical methods to mutually verify and complement the characteristics of the research object. Ore body extension trend inversion information can be a collection of information obtained through joint inversion analysis, accurately characterizing the extension characteristics and distribution patterns of ion-adsorption type rare earth ore bodies in underground three-dimensional space.

[0034] Specifically, in the analysis of rare earth ore body extension trends, the formation and distribution of ore bodies are highly correlated with the geological environment. Relying solely on sensor data analysis will deviate from the geological background, leading to large deviations in the inversion results. By first acquiring and processing the geological environment information set, and then constructing a geological-sensor data coupled inversion model, the two types of data are integrated to carry out joint inversion, so as to achieve mutual verification between geological and measured data and obtain ore body extension trend inversion information that is consistent with reality.

[0035] S203. Based on the inversion information of the ore body extension trend, adaptive analysis is performed on the abnormal sections of the underground ore body occurrence, and a rare earth underground ore body occurrence analysis log containing the three-dimensional results of the ore body occurrence is generated and output.

[0036] The occurrence of an underground ore body can be its specific morphology, spatial location, boundaries, strike, dip, dip angle, and thickness variations in three-dimensional space, including geometric and spatial attributes. Anomaly zones can be localized spatial areas that significantly deviate from the overall regularity revealed by the ore body's extension trend inversion information. Three-dimensional ore body occurrence results can be digitally generated results that fully describe the occurrence of the underground ore body, expressed as a high-precision three-dimensional raster model or surface model. A rare earth underground ore body occurrence analysis log can be a complete and traceable technical report recording the entire analysis process.

[0037] Specifically, in the final analysis of ore body occurrence, the identification and analysis of abnormal sections is the key to safe and efficient mining. Existing fixed parameter analysis has low accuracy and poor flexibility, and is prone to omissions and misjudgments. By extracting inversion information features, machine learning combined with spatial analysis methods is used to achieve accurate identification and quantification of abnormal sections, and then three-dimensional results are constructed. Finally, the information of the whole process is integrated to generate and output standardized analysis logs.

[0038] The method provided in this embodiment solves the core problem of spatiotemporal heterogeneity of multi-source sensor data in rare earth exploration, lays a standardized data foundation for ore body analysis, and makes the analysis of ore body extension trend more accurate and comprehensive by combining geological environmental information with joint inversion. Adaptive analysis enables intelligent identification of abnormal sections, and three-dimensional results make the occurrence of ore bodies more intuitive. The output analysis log provides a scientific basis for rare earth exploration and mining decisions, greatly improves analysis efficiency and accuracy, promotes the intelligent and precise development of rare earth exploration and development, and helps to develop rare earth resources efficiently and rationally.

[0039] In some embodiments, the multi-source heterogeneous sensing dataset includes microseismic response data and electromagnetic detection data. Based on the microseismic response data, discrete event timestamps are analyzed, and synchronization alignment is performed by introducing a unified time reference event to obtain time-unified microseismic data. Based on the electromagnetic detection data, the coordinates of multi-source measuring points are analyzed, and scale normalization is performed through a unified grid spatial resampling rule to obtain spatial grid electromagnetic data. Based on the time-unified microseismic data and combined with the spatial grid electromagnetic data, the spatial coordinates and temporal attributes of the two are registered and fused to obtain spatiotemporally aligned multi-source sensing data.

[0040] Microseismic response data can be discrete event sensing data reflecting micro-fractures in the rock mass of a rare earth ore body, acquired through microseismic monitoring equipment. Electromagnetic detection data can be spatial distribution sensing data characterizing the electromagnetic properties of the medium in a rare earth ore body, obtained through electromagnetic detection equipment. A unified time reference event can be a landmark geological or artificial event with a clear time node selected for calibrating the timestamps of discrete microseismic events. Time-unified microseismic data can be microseismic data obtained by synchronizing and aligning the discrete event timestamps of microseismic response data to a unified time reference. Multi-source measuring point coordinates can be the original spatial location coordinate information corresponding to each measuring point during electromagnetic detection. A unified grid spatial resampling rule can be a unified spatial grid division and data resampling criterion for electromagnetic data scale normalization, formulated based on the exploration needs of the mining area. Spatial grid electromagnetic data can be electromagnetic data obtained by scaling electromagnetic detection data according to the unified grid spatial resampling rule. Registration and fusion analysis can be an analytical process of matching and verifying the spatial coordinates and temporal attributes of time-unified microseismic data and spatial grid electromagnetic data, and fusing features.

[0041] Specifically, in the multi-source sensor data processing for rare earth orebody occurrence analysis, data exhibit biases in sampling scale, timestamps, and spatial location. Direct fusion of these data can lead to data feature distortion, causing misjudgments in orebody trend inversion and even resource location deviations in mining area exploration. By unifying the spatiotemporal scales of microseismic and electromagnetic data, various biases from multi-source heterogeneous data are eliminated, providing reliable basic data support for subsequent joint inversion of orebody extension trends and avoiding analytical errors caused by data biases.

[0042] In the specific analysis process: First, time synchronization is performed: the original timestamps of all microseismic events are read, and the 1PPS (pulses per second) signal provided by the GPS receiver is introduced as a unified time reference event. By calculating the difference between the event time recorded by each acquisition station and the 1PPS signal time received by the station at the same time (i.e., clock offset), the timestamps of all microseismic events are compensated and corrected to generate unified microseismic data with all event times consistent with the GPS time scale. Second, spatial normalization is performed: the three-dimensional coordinates (such as geodetic coordinates or local coordinates) and their apparent resistivity and other attribute values ​​of all electromagnetic measurement points are read. A three-dimensional regular grid system covering the target mining area is defined as a unified grid spatial resampling rule. For example, the grid origin is the southwest corner of the mining area, the grid spacing in the X and Y directions is 10 meters, and the spacing in the Z direction (depth) is 5 meters. Using the bilinear interpolation algorithm, the attribute values ​​of each irregular measurement point are interpolated and assigned to the center of the regular grid cell where it is located and its surroundings, generating spatial grid electromagnetic data in which each grid cell contains electromagnetic attribute values. Finally, registration and fusion are performed: the system finds and maps each microseismic event after time unification to the corresponding three-dimensional grid cell in the spatial grid electromagnetic data according to its three-dimensional coordinates (X,Y,Z). At the same time, the data acquisition time interval of the electromagnetic grid cell is extracted. For microseismic events falling within the same electromagnetic detection time interval, they are bound to the electromagnetic response characteristics (such as apparent resistivity) of the grid cell to form a "spatiotemporal correlated event pair". All such event pairs constitute spatiotemporally aligned multi-source sensing data that can be directly used for subsequent analysis.

[0043] In alternative or modified implementations: the time reference event is not limited to GPS 1PPS, but can also use the server time synchronized by the high-precision network clock protocol (NTP) as the reference. The spatial resampling rule is not limited to the regular grid with equal spacing, but can also use the unstructured grid based on the stratigraphic model. In this case, the resampling algorithm needs to be changed to Kriging interpolation or natural neighbor interpolation. For time window association matching, in addition to fixed time interval matching, sliding time windows or event-triggered dynamic time windows can also be used for association to adapt to the rhythm of different exploration tasks.

[0044] In some embodiments, based on time-unified microseismic data, the spatial coordinate distribution of microseismic events is analyzed. Combined with the unified grid space followed by the spatial grid electromagnetic data, the coordinates of each microseismic event are mapped to the corresponding spatial grid cell, resulting in a spatially registered microseismic event sequence. Based on the microseismic event sequence, the unified time reference corresponding to each event is analyzed. Combined with the detection time interval of the spatial grid electromagnetic data on the corresponding grid cell, time window correlation matching is performed between the microseismic events and the electromagnetic detection results, resulting in a set of spatiotemporally correlated event pairs. Based on the set of spatiotemporally correlated event pairs, the correspondence between microseismic event attributes and electromagnetic response characteristics within the same grid cell is analyzed. By superimposing and verifying the consistency of the two characteristics, a feature description fusing microseismic and electromagnetic information within each grid cell is generated, resulting in spatiotemporally aligned multi-source sensing data.

[0045] A microseismic event sequence can be a set of microseismic events arranged in temporal or spatial order after their coordinates are mapped to corresponding spatial grid cells. A unified time reference can be a unified time standard set for the time alignment of microseismic data. A detection time interval can be the time range within which electromagnetic data from a spatial grid is detected on the corresponding grid cell. Time window correlation matching is the process of matching the time information of microseismic events with the electromagnetic detection time interval of the corresponding grid cell. A spatiotemporally correlated event pair set can be a set of event pairs consisting of successfully matched microseismic events and electromagnetic detection results within a time window. Microseismic event attributes can be characteristic information of microseismic events, such as energy, magnitude, and occurrence time. Electromagnetic response characteristics can be information reflecting geological features obtained from electromagnetic detection, such as apparent resistivity and response amplitude. Consistency verification can be a verification method for confirming the logical correlation between microseismic event attributes and electromagnetic response characteristics within the same grid cell.

[0046] Specifically, in the multi-source data fusion process for rare earth orebody extension trend inversion, the lack of spatiotemporal registration between microseismic and electromagnetic data leads to data misalignment and confused feature correspondences, directly causing deviations in the identification of orebody anomaly zones. Subsequent inversion analysis loses reliable data support and cannot accurately delineate the orebody extension boundaries. By accurately registering and fusing microseismic and electromagnetic data in both spatiotemporal dimensions, the problem of data misalignment is effectively solved, allowing the two types of data features to correspond precisely. This provides consistent and reliable fused sensor data for subsequent orebody inversion, laying a solid foundation for the identification of orebody anomaly zones and the analysis of extension trends.

[0047] In the specific analysis process: First, spatial location registration: each event record in the "time-unified microseismic data" is read and its three-dimensional coordinates (X,Y,Z) are extracted. At the same time, the predefined three-dimensional grid index file of the "spatial grid electromagnetic data" is loaded. This file defines the center coordinates and boundaries of each grid. The "nearest neighbor grid" algorithm is used to calculate the Euclidean distance from the coordinates of each microseismic event to the center of all grids, assign it to the nearest grid cell, and record the globally unique ID of the grid. All events that have been assigned are sorted by time to form a "spatial location registered microseismic event sequence". Each event in the sequence is appended with its grid ID. Next, time window association matching: For each microseismic event in the sequence, the system uses its occurrence timestamp t_event as the center, extending forward and backward by a preset duration Δt (e.g., ±30 minutes), forming a time window [t_event-Δt, t_event+Δt]. Then, based on the grid ID attached to the event, it retrieves all data blocks corresponding to this grid in the "spatial grid electromagnetic data" and checks the acquisition time interval [t_em_start, t_em_end] of each data block. If [t_em_start, t_em_end] intersects with [t_event-Δt, t_event+Δt] (i.e., the time windows overlap), then this microseismic event is considered to be spatiotemporally associated with this segment of electromagnetic data, forming a "spatiotemporally associated event pair". After traversing all microseismic events, a complete "set of spatiotemporally associated event pairs" is obtained. Finally, feature fusion and verification: For each event pair in the set, from the microseismic... The data extracts attributes such as energy and dominant frequency of the event, and extracts features such as apparent resistivity and phase from the corresponding electromagnetic data block. A new data structure corresponding to the grid cell is created in memory. The fusion algorithm adopts a weighted superposition method. For example, the normalized microseismic energy attenuation index and the electromagnetic apparent resistivity anomaly coefficient are weighted and summed to generate an initial value of "cooperative anomaly intensity". Then, "consistency verification" is performed: it checks whether the microseismic energy has significantly attenuated (below the background threshold) and whether the electromagnetic apparent resistivity has decreased synchronously (below the background threshold). If the changes in both are in the same direction (both indicate possible mineralization), the verification is passed, and the "cooperative anomaly intensity" value is retained and may be enhanced. If the directions are opposite or one has no significant change, the verification fails, and its weight is significantly reduced or it is marked as unreliable data. After this operation is completed for all grid cells, a "spatiotemporally aligned multi-source sensor data" cube with a fused feature vector (including cooperative anomaly intensity, verification flag, etc.) for each grid cell is output.

[0048] In alternative or modified implementations: Spatial registration can employ a more refined "inverse distance weighting" method instead of the nearest neighbor principle. This means that a microseismic event can simultaneously affect multiple surrounding grids, with the influence weight calculated inversely proportional to the distance from the event to the center of each grid. This more smoothly reflects the contribution of the event's point source characteristics to the surrounding grids. The time window Δt can be dynamically adjusted based on the magnitude or energy of the microseismic event. Events with higher energy may represent more persistent or wider-ranging rupture processes, thus allowing for the association of longer time windows. Consistency verification can be based on a machine learning model, using historical data from known mining areas to train a classifier. The input is a combination vector of microseismic and electromagnetic features within a grid, and the output is the probability of the combination indicating a mineralization anomaly. This replaces the fixed threshold judgment logic, making the verification more adaptive and robust.

[0049] In some embodiments, the geological environment information set includes hydrogeological information and ore body weathering crust information. Based on the hydrogeological information, the constraint of groundwater runoff on the ion migration and enrichment law of the ore body is analyzed to obtain groundwater constraint information. Based on the ore body weathering crust information, the constraint of weathering crust stratification on the ore body basement morphology and three-dimensional distribution framework is analyzed to obtain weathering crust morphology information. The groundwater constraint information and weathering crust morphology information are integrated to construct the basic geological spatial constraint framework of the ore body. Based on spatiotemporally aligned multi-source sensing data, the interaction between microseismic event attributes and electromagnetic response characteristics is analyzed. According to the synergistic change information of spatial attenuation of microseismic energy and abnormal response of electromagnetic apparent resistivity in the ore body occurrence area, a multi-source synergistic anomaly feature set is obtained. Based on the basic geological spatial constraint framework and combined with the multi-source synergistic anomaly feature set, the coupling relationship between geological constraints and geophysical anomalies in three-dimensional space is analyzed to optimize and delineate the extension boundary and trend of the ore body, and to obtain ore body extension trend inversion information.

[0050] Hydrogeological information can be geological data reflecting the distribution, movement patterns, and hydrodynamic characteristics of groundwater in the mining area, providing hydrological basis for analyzing ion migration and enrichment. Ore body weathering crust information can be geological data characterizing the layered structure, basement morphology, and distribution characteristics of the weathering crust of ion-adsorption type rare earth ore bodies. Groundwater constraint information can be data derived from hydrogeological information analysis, showing the limitations and influences of groundwater runoff on the formation of ion migration and enrichment patterns in the ore body. Weathering crust morphology information can be data derived from weathering crust information analysis, showing the constraint characteristics of weathering crust layering on the basement morphology and three-dimensional distribution framework of the ore body. The basic geological spatial constraint framework can be a three-dimensional spatial constraint model that integrates groundwater constraint information and weathering crust morphology information to delineate geological boundaries for ore body inversion. The multi-source synergistic anomaly feature set can be a set of synergistic variation characteristic data of extracted microseismic energy spatial attenuation and electromagnetic apparent resistivity anomalies in the ore body occurrence area.

[0051] Specifically, in the joint inversion of the extension trend of ion-adsorption rare earth ore bodies, relying solely on geophysical data without geological constraints can easily lead to distorted inversion boundaries, mistakenly including non-occurring areas. Furthermore, biases in single-type geophysical characteristics can cause the inferred ore body extension trend to deviate significantly from actual geological conditions. By constructing a dedicated spatial constraint framework using geological environmental information, a reasonable range for ore body inversion is defined. Combined with multi-source geophysical collaborative anomaly feature analysis, the ore body extension trend inversion results are highly consistent with geological reality, significantly improving the accuracy and geological conformity of the inversion.

[0052] In the specific analysis process: First, hydrogeological information is processed: isotope maps and seepage field simulation results obtained from exploration are imported. The program automatically extracts the main flow direction and divides the catchment units. The blocking or guiding effect of each unit on rare earth ion migration is quantified into the "constraint strength value" of that unit, generating groundwater constraint information (a three-dimensional scalar field). Simultaneously, weathering crust information is processed: based on boreholes and geological profiles, a three-dimensional surface model of the weathering crust interface is established. The program calculates the normal vector at each point of this surface to indicate the basement dip, and assigns different "mineralization probabilities" to different strata based on the thickness of the weathering crust layers. The system generates weathering crust morphology information (a composite model containing geometric surfaces and attribute weights) by assigning weights to each grid cell. Then, these two weights are superimposed: in the three-dimensional grid space, each grid cell is simultaneously assigned an "intensity value" from groundwater constraints and a "probability weight" from the weathering crust morphology. These weights are then fused through a weighted product to form a basic geological spatial constraint framework—a three-dimensional probability volume. High-value regions represent the geologically most favorable mineralization spaces. Next, a multi-source collaborative anomaly feature set is constructed: for each spatiotemporally aligned grid cell, the gradient of energy decay of its internal microseismic events with propagation distance (d) is calculated. The program extracts the electromagnetic apparent resistivity anomalous amplitude (Δρ) of the cell relative to the background field (B / m). A collaborative discrimination function is set; for example, when both "energy decay gradient > threshold G and apparent resistivity anomalous amplitude < threshold R" are satisfied, the cell is marked as a "collaborative anomalous cell". After traversing all grids, spatial clustering is performed on these anomalous cells, and isolated points are removed, ultimately forming a multi-source collaborative anomalous feature set composed of continuous anomalous clusters. Finally, optimization and delineation are performed: the collaborative anomalous clusters are projected onto the basic geological spatial constraint framework (probabilistic volume), and the program calculates the anomalous cluster value for each cluster. The average probability of geological constraints at the location of the block is set, and a probability threshold is set. Only anomalous blocks with a geological constraint probability higher than the threshold are retained for subsequent delineation. For these blocks with "geological-geophysical dual verification", an implicit function surface fitting algorithm (such as radial basis function network) is used to generate a preliminary ore body wrapping surface using its three-dimensional spatial coordinates. Then, using this wrapping surface as the initial input, geological smoothing constraints (such as the ore body shape usually changing continuously and gradually) are introduced for iterative optimization. Finally, smooth, continuous and geologically consistent ore body extension trend inversion information (i.e., the optimized three-dimensional ore body model) is output.

[0053] In alternative or modified implementations: When constructing the basic geological spatial constraint framework, if weighted product fusion is not used, fuzzy logic rules can be used instead. For example, define "IF groundwater constraint strength is high AND weathering crust mineralization probability is high THEN comprehensive constraint level is 'extremely high'". For the construction of multi-source collaborative anomaly feature sets, in addition to preset threshold discrimination, unsupervised machine learning algorithms (such as isolated forests) can be used to automatically detect outliers in the joint space of microseismic and electromagnetic features as collaborative anomalies. In the optimization and deduction stage, the implicit function surface fitting can be replaced by a method based on three-dimensional kriging interpolation or neural network to directly regress the ore body boundary. Geological smoothing constraints can also be replaced by an anisotropic diffusion smoothing algorithm based on partial differential equations to adapt to ore bodies with different occurrences.

[0054] In some embodiments, based on time-unified microseismic data, the energy attenuation gradient characteristics of microseismic event attributes during spatial propagation within each spatial grid cell are analyzed to obtain microseismic energy attenuation information; based on spatial grid electromagnetic data, the apparent resistivity variation characteristics of electromagnetic response features within the same spatial grid cell in three-dimensional space are analyzed to obtain electromagnetic apparent resistivity anomaly information; based on the microseismic energy attenuation information and combined with the electromagnetic apparent resistivity anomaly information, the synchronicity of their changes within the same spatial grid cell is analyzed, and based on the regions where the increase in microseismic energy attenuation gradient and the decrease in electromagnetic apparent resistivity anomaly occur simultaneously in space, orebody-related co-anomaly zones are identified; based on the co-anomaly zones, the corresponding intensity and spatial continuity of microseismic energy attenuation characteristics and electromagnetic apparent resistivity anomaly characteristics are analyzed, and a multi-source co-anomaly feature set is obtained through joint measurement and spatial aggregation of the two characteristics.

[0055] Microseismic energy attenuation gradient characteristics can be defined as the attenuation rate and gradient changes of microseismic event energy during spatial propagation, varying with geological media and propagation distance. Microseismic energy attenuation information can be a dataset composed of microseismic energy attenuation gradient characteristics from each spatial grid cell, reflecting the spatial propagation law of microseismic energy. Electromagnetic apparent resistivity anomaly information can be characteristic data of electromagnetic apparent resistivity deviations from the normal geological background value within each spatial grid cell, reflecting electrical anomalies in the subsurface medium. Co-anomaly zones can be areas within the same spatial grid cell where an increase in microseismic energy attenuation gradient and a decrease in electromagnetic apparent resistivity anomaly occur simultaneously, and are related to ore body occurrence. Joint measurement can be an analytical method for quantitatively evaluating the corresponding intensity of microseismic energy attenuation characteristics and electromagnetic apparent resistivity anomaly characteristics within co-anomaly zones. Spatial aggregation can be a processing method that clusters and integrates the features within co-anomaly zones that have undergone joint measurement according to their spatial location relationships.

[0056] Specifically, in the process of multi-source data fusion for ore body extension trend inversion, analyzing microseismic or electromagnetic data alone is prone to false anomalies, unable to match the geophysical exploration collaborative characteristics of the ore body occurrence, which will lead to deviation in the identification of abnormal areas and directly affect the accuracy of subsequent ore body boundary delineation. Through the collaborative analysis of microseismic and electromagnetic data, accurately identify the collaborative abnormal areas related to the ore body occurrence, effectively eliminate the interfering false anomalies of single data, and provide accurate and reliable multi-source collaborative anomaly feature basis for ore body extension trend inversion.

[0057] In the specific analysis process: First, based on the time-unified microseismic data, for each microseismic event in each grid cell, use the formula gradient = (E_near - E_far) / distance to calculate the attenuation gradient of its energy from the near field to the far field, where E_near and E_far are the energy values sorted according to the distance between the event and the sensor, and distance is the corresponding sensor spacing. After traversing all valid events, take the statistical average (such as the median) of the gradients in the cell as the "microseismic energy attenuation information" value of the cell. Second, extract the apparent resistivity value of each grid cell from the spatial grid electromagnetic data, and use the method of "mean ± 2 times standard deviation" or regional background field fitting residual method to identify the grids below the normal threshold (for example, 30% lower than the regional background value). After normalizing the relative difference between its apparent resistivity value and the background value, it is used as the "electromagnetic apparent resistivity anomaly information" value of the cell. Then, align the two data sets according to the spatial grid index, and calculate the Pearson correlation coefficient or trend synchronization score of the two information values in each cell (for example, define that the attenuation gradient increases and the apparent resistivity decreases as "positive synchronization"). Next, set double thresholds for identification: the first is the single anomaly threshold (such as attenuation gradient > X, apparent resistivity anomaly value < Y), and the second is the synchronization score threshold (such as correlation coefficient > 0.5). Only when a certain grid cell meets both the single anomaly and synchronization requirements at the same time, it is marked as a "collaborative abnormal area". Finally, for each "collaborative abnormal area", calculate its "collaborative intensity" feature (such as the weighted product of the two anomaly values or the fusion value based on the synchronization score) and "spatial continuity" feature (such as the number of adjacent collaborative abnormal cells to this cell). Combine these features (location, collaborative intensity, continuity, etc.) of all collaborative abnormal areas into a structured list, that is, generate a multi-source collaborative anomaly feature set.

[0058] In alternative or modified implementations: when calculating synchronicity, it is not limited to the Pearson correlation coefficient. Mutual information can be used to measure the statistical dependence between the two, which is more robust to nonlinear relationships. When constructing the feature set, in addition to the above features, "abnormal morphological features" (such as calculating the spatial distribution shape of the co-anomaly zone in a local range) or "depth-weighted features" (because ore bodies are usually located at a specific weathering crust depth) can also be introduced. In addition, the threshold for identifying "co-anomaly zones" can be determined using an adaptive method. For example, the Otsu algorithm can be used to calculate the binarized threshold for the attenuation gradient data and the apparent resistivity anomaly data respectively, and then the synchronicity requirement can be combined for judgment.

[0059] In some embodiments, based on microseismic energy attenuation information, the gradient change characteristics along the depth direction within each spatial grid cell are analyzed to obtain microseismic gradient change information; based on electromagnetic apparent resistivity anomaly information, the anomalous fluctuation characteristics with increasing depth within the same spatial grid cell are analyzed to obtain electromagnetic anomalous fluctuation information; based on the microseismic gradient change information and combined with the electromagnetic anomalous fluctuation information, the degree of matching between the two in the corresponding spatial grid cell and depth interval is analyzed, and synchronous change regions are identified according to the degree of synergy between gradient increase and anomalous decrease in spatial location and change pattern; based on the synchronous change regions, the continuity and correlation strength of microseismic gradient and electromagnetic anomalous fluctuation in three-dimensional space are analyzed to obtain the coupling relationship.

[0060] Microseismic gradient change information can be obtained by analyzing the gradient change characteristics along the depth direction within each spatial grid cell based on microseismic energy attenuation information. Electromagnetic anomaly fluctuation information can be obtained by analyzing the anomalous fluctuation characteristics with increasing depth within the same spatial grid cell based on electromagnetic apparent resistivity anomaly information. Synchronous change regions can be areas in the ore body region where the changing trends of increasing microseismic gradient and decreasing electromagnetic anomaly are highly matched in terms of spatial location and change pattern.

[0061] Specifically, in analyzing the three-dimensional coupling relationship between geological constraints and geophysical anomalies, if the microseismic and electromagnetic data are not matched in depth, misjudgments of anomaly associations due to single data biases can easily lead to significant errors in subsequent orebody boundary extrapolation and delineation. To address this issue, precise collaborative analysis of microseismic and electromagnetic geophysical data in the depth dimension accurately quantifies their three-dimensional coupling relationship, providing a reliable basis for the correlation analysis of geological constraints and geophysical anomalies, and improving the accuracy of subsequent orebody extension trend extrapolation.

[0062] In the specific analysis process: First, from the constructed 3D data volume of "microseismic energy attenuation information" indexed by a unified grid, for each grid cell (such as the cell numbered G(i,j,k)), extract the attenuation value sequences of its adjacent cells above and below along the depth direction (k direction), perform a first-order difference calculation on this sequence, and obtain the "microseismic energy attenuation gradient" value of this grid cell in depth. After traversing all cells, form the 3D data volume G_grad of "microseismic gradient change information". Second, from the 3D data volume R_anom of "electromagnetic apparent resistivity anomaly information", for each grid cell G(i,j,k), extract the apparent resistivity anomaly value sequence within the local window centered on it along the depth direction (such as ±2 depth cells). Calculate the standard deviation of this sequence or the sum of squared residuals after detrending as the "electromagnetic anomaly fluctuation intensity" value of this cell, and form the 3D data volume F_fluct of "electromagnetic anomaly fluctuation information". Then, perform a co-analysis: For each vertical column of grid cells with the same horizontal position (i,j), measure the synchrony of the depth profile curves of G_grad and F_fluct in this column. Specifically, use the sliding window dynamic time warping (DTW) algorithm to calculate the matching distance D_match(i,j) in the morphology of the two curves. Set a threshold Th_sync. If D_match(i,j) < Th_sync, then consider the horizontal area where this vertical column is located as the "synchronously changing area" and record its horizontal range. Finally, quantify the coupling relationship: In all the marked "synchronously changing areas", calculate the Pearson correlation coefficient ρ(i,j) of the two profile curves of G_grad and F_fluct in each vertical column, and count the connected volume V_coupled and the average correlation coefficient ρ_avg of these high correlation coefficient areas in 3D space. Finally, the "coupling relationship" is quantified as a comprehensive description set including the spatial range of the "synchronously changing area", the average correlation coefficient ρ_avg, and the spatial connectivity index (such as V_coupled).

[0063] In alternative or modified implementations: the specific algorithm for the above-mentioned collaborative analysis can be replaced. For example, the matching degree calculation can be performed without using DTW, instead using the proportion of consistency of the first derivative signs (increase / decrease) of the two depth curves at corresponding depth points. When the consistency proportion exceeds a preset value (e.g., 75%), it is determined to be synchronous. For the quantification of "coupling relationship", in addition to using the correlation coefficient, in alternative implementations, a variogram model based on geostatistics can be introduced to calculate the ranges of G_grad and F_fluct in three spatial directions respectively. If the ranges of the two are similar and have high consistency in the main mineralization direction, they are considered to be strongly coupled. Another modification is to combine the identification of "synchronous change areas" with cluster analysis, considering not only the matching degree of a single vertical column, but also its spatial clustering on the horizontal plane. Spatial clustering algorithms such as DBSCAN are used to aggregate discrete high-matching points into patches to more robustly determine the boundaries of "synchronous change areas".

[0064] In some embodiments, based on synchronously changing regions, the stability and consistency of microseismic gradient change information and electromagnetic anomaly fluctuation information in space are analyzed to obtain mineralization clue regions. According to the coupling relationship, the continuity and change trend of the mineralization clue regions in three-dimensional space are analyzed. Combined with the basic geological spatial constraint framework, the main extension direction of the ore body in the horizontal and vertical directions is inferred to obtain the three-dimensional morphology inference information of the ore body. Based on the three-dimensional morphology inference information of the ore body and the multi-source collaborative anomaly feature set, within the boundary of the basic geological spatial constraint framework, the spatial contour of the ore body is smoothly connected and the boundary is optimized to delineate the continuous distribution range of the ore body in three-dimensional space and obtain the ore body extension trend inversion information.

[0065] Mineralization clue regions can be areas where the spatial stability and consistency of microseismic gradient changes and electromagnetic anomalies meet preset standards within synchronously changing regions. The coupling relationship can be the continuity and correlation strength characteristics of microseismic gradients and electromagnetic anomalies in three-dimensional space. The three-dimensional morphology deduction information of the ore body can be information related to the horizontal and vertical main extension directions of the ore body derived from the mineralization clue regions, combined with a geological constraint framework.

[0066] Specifically, in the process of delineating the extension boundaries and trends of ore bodies, directly delineating them based on coupling relationships can easily lead to problems such as faults in mineralized areas and blurred boundaries. Furthermore, deviating from the geological framework can cause the deduction results to deviate from reality, resulting in misjudgments of exploration target areas and a significant reduction in the efficiency of rare earth resource exploration. To address these issues, by screening mineralized clue areas and combining them with the geological framework to deduce the ore body morphology, and then optimizing the outline to delineate the boundaries, the inversion results of the ore body extension trend are made more consistent with the actual occurrence state, accurately delineating the continuous distribution range, defining precise target areas for rare earth mineral exploration, and improving the scientific nature of exploration.

[0067] In the specific analysis process: First, the "synchronous change area" and its corresponding "coupling relationship" intensity value are treated as discrete three-dimensional spatial data points (each point contains spatial coordinates XYZ, microseismic gradient value G, electromagnetic anomaly value E, and coupling strength C). Using three-dimensional geological modeling software (such as Leapfrog Geo, GOCAD) or a custom algorithm, spatial interpolation (such as inverse distance weighting, kriging) is performed on these data points to generate a preliminary three-dimensional volume model of the "mineralization clue area." The attribute value of each volume element reflects the comprehensive intensity of the mineralization clue at that location. Next, based on this three-dimensional volume model, principal component analysis (PCA) or directional statistics is used to analyze the aggregation direction and extension trend of high attribute value volume elements in space. Combined with the known stratigraphic attitude and structural line direction in the "basic geological spatial constraint framework," the "main extension direction" of the ore body in the horizontal and vertical directions is inferred, forming preliminary "ore body three-dimensional morphology inference information." For example, it is inferred that the ore body mainly extends deep along the 30° northeast direction with a dip angle of approximately 45°. Finally, within the "basic geological spatial constraint..." Within the boundaries defined by the "framework" (such as the bottom boundary of the weathered crust and fault boundaries), the derived three-dimensional morphology serves as the skeleton. The anomaly intensity at each spatial location in the "multi-source collaborative anomaly feature set" is used as the control weight to perform "smoothing connection and boundary optimization" on the preliminary skeleton. Specifically, in the modeling software, the surface of the ore body is meshed and smoothed using moving average filtering or Kriging interpolation with geological trend surfaces, with anomaly intensity as a constraint. For the boundaries, anomaly intensity threshold is set (for example, areas with collaborative anomaly feature values ​​below a certain threshold are removed), and morphological algorithms such as erosion-dilation are used for optimization. Finally, a continuous, smooth, and geologically consistent three-dimensional "continuous distribution range" of the ore body is delineated, which is the "ore body extension trend inversion information".

[0068] In alternative or modified implementations: the generation of "mineralization clue regions" can be achieved without relying on commercial software, using machine learning-based classification algorithms (such as random forests and support vector machines). A model is trained using the multidimensional features (coordinates, gradient values, outliers, coupling strength, etc.) of the "synchronously changing regions" as input to automatically identify and cluster high-probability mineralization spatial units. The inference of the "main extension direction" can also employ three-dimensional Hough transform to detect dominant planes or linear trends in the spatial point cloud. For "smooth connectivity and boundary optimization," in addition to interpolation methods, implicit modeling techniques can be used to automatically generate a smooth orebody model by solving a constrained scalar field function (whose isosurface is the orebody boundary); or an active contour model (Snake model) can be used, allowing the initial contour to evolve under the influence of the anomaly feature field and the geological constraint framework until it converges to the optimal boundary.

[0069] In some embodiments, based on the ore body extension trend inversion information, the local abrupt changes and discontinuities in morphology, thickness and geophysical response characteristics within the three-dimensional distribution range of the ore body are analyzed to obtain ore body occurrence anomaly segment information; based on the ore body occurrence anomaly segment information, combined with the basic geological spatial constraint framework, and according to the adaptive weighted correction of the anomaly cause, the ore body occurrence information is obtained; integrating the ore body extension trend inversion information, ore body occurrence anomaly segment information and the corresponding ore body occurrence information, a rare earth underground ore body occurrence analysis log is generated and output.

[0070] Adaptive analysis can be an analytical process that automatically matches and analyzes different characteristics of anomalous orebody occurrence segments to identify anomalies, determine their causes, and perform differentiated corrections. Information on anomalous orebody occurrence segments can be comprehensive data recording the location, extent, degree of morphological abrupt changes, thickness variation rate, and geophysical response instability characteristics of the anomalous segments. Adaptive weighted correction of anomaly causes can be a process of assigning dynamic weights to different causes based on the specific geological genesis of the anomalous segments, and differentially correcting the occurrence projection results. Orebody occurrence information, after anomaly correction, can be the final orebody occurrence data that accurately reflects the three-dimensional morphology, thickness, and distribution characteristics of the orebody as a whole and the anomalous segments.

[0071] Specifically, in the process of analyzing and outputting the occurrence of rare earth underground ore bodies, if adaptive analysis and correction of anomalous sections are not performed, the occurrence results will not match the actual geological conditions. This directly leads to deviations in the delineation of exploration target areas and errors in mining planning, as well as wasted costs and resource losses in rare earth resource exploration and mining. To address these issues, adaptive analysis and precise correction of anomalous ore body occurrence sections ensure that the output three-dimensional ore body occurrence results are more closely aligned with actual geological conditions. This provides accurate data support for the delineation of rare earth exploration target areas and the formulation of mining plans, significantly improving the scientific rigor and efficiency of exploration and mining.

[0072] In the specific analysis process: First, based on the inversion information of the ore body extension trend (i.e., a three-dimensional volume model), each voxel unit is traversed programmatically to calculate its morphological gradient (such as curvature change) and thickness change rate with adjacent units, and the corresponding microseismic energy attenuation gradient and electromagnetic apparent resistivity anomaly value are retrieved simultaneously to jointly constitute "occurrence anomaly characteristic information". Then, this characteristic information is spatially superimposed with hydrogeological maps (such as the distribution of water-rich areas) and weathering crust layering models: For example, if an anomaly area shows a high morphological gradient, is located in a known active groundwater zone, and has drastic fluctuations in electromagnetic response, then its main cause is determined to be "the influence of hydrological activity"; if the anomaly area coincides with the location of the abrupt change in the weathering crust basement and the microseismic signal is chaotic, then it is determined to be "weathering crust interface change". Then, a weight configuration table is built in to preset different correction intensity coefficients (weights) for different causes (hydrological, interface, tectonic). For example, in the "hydrological activity impact" zone, which may cause irregular dissolution of the ore body boundary, the correction weights tend to be based on the local boundary contraction according to electromagnetic anomaly fluctuations; in the "weathering crust interface change" zone, the correction weights tend to be based on the microseismic gradient change to make the ore body shape conform to the basement interface. Finally, these differentiated weights are used to locally move and adjust the shape of the initially inferred ore body boundary to generate "corrected ore body occurrence information", and to ensure that it does not break through the basic geological spatial constraint framework as a whole, and is logically consistent with the spatial distribution of the multi-source collaborative anomaly feature set. The final coordinated "ore body occurrence information" is obtained through iterative feedback verification, and all intermediate and final results are integrated to generate a structured analysis log.

[0073] In alternative or modified implementations: the extraction of occurrence anomaly features is not limited to morphological gradients and rates of change; more complex geometric and physical property indicators such as texture analysis and anisotropy indices can be incorporated. For the determination of anomaly causes, machine learning classification models (such as random forests) can be introduced, trained using a large number of samples from explored areas (anomaly features + verified geological causes), thereby achieving intelligent and rapid determination of the causes of anomaly segments in new areas. Furthermore, the adaptive weighted correction algorithm is not limited to a preset weight table; a fuzzy logic system can be used to dynamically calculate a correction confidence level between 0 and 1 based on input variables such as the strength of anomaly features and the credibility of the cause, thereby flexibly adjusting the inference results and making the correction process more adaptive and robust.

[0074] In some embodiments, based on the information of orebody occurrence anomaly sections, the characteristics of each anomaly section in terms of morphological abrupt change, thickness variation rate, and geophysical response instability are analyzed to obtain occurrence anomaly characteristic information. Based on the occurrence anomaly characteristic information, combined with hydrogeological information and orebody weathering crust information, the correlation between different anomaly characteristics and geological environmental genesis is analyzed to obtain anomaly gene determination information. Based on the anomaly gene determination information, combined with orebody three-dimensional morphological inference information, the occurrence inference results of the anomaly sections are differentially weighted and corrected according to the influence weights of different genesis such as hydrological activity, weathering crust interface changes, and tectonic disturbances on the continuity of orebody occurrence, to obtain corrected orebody occurrence information. Based on the corrected orebody occurrence information, combined with the basic geological spatial constraint framework, the coordination between the corrected results and the overall geological framework is analyzed. Through feedback verification with the multi-source collaborative anomaly feature set, the final coordinated orebody occurrence information is obtained.

[0075] Anomaly cause identification information can be obtained by combining hydrogeological and ore body weathering crust information, and analyzing the correlation between different occurrence anomaly characteristics and geological environment genesis. Weathering crust interface changes can refer to alterations in the location and morphology of interfaces between different layers of the ore body's weathering crust, affecting the ore body's occurrence. Tectonic disturbances can be disturbances to the continuity of ore body occurrence caused by rock mass deformation and displacement resulting from tectonic movements. Influence weights can be quantitative values ​​determined based on the different causes of hydrological activity, weathering crust interface changes, and tectonic disturbances, representing the degree of influence on the continuity of ore body occurrence. Corrected ore body occurrence information can be obtained by differentially weighting and correcting the occurrence projection results of the anomaly section based on anomaly cause identification information and influence weights.

[0076] Specifically, in the analysis of orebody occurrence anomaly zones, directly correcting without distinguishing the causes of the anomalies can lead to discrepancies between the occurrence projection results and the actual geology. It can also distort the three-dimensional distribution contour of the orebody, affecting the accuracy of exploration borehole layout and resource reserve calculations, resulting in exploration decision-making errors and resource waste. To address these issues, by accurately identifying the causes of the anomalies and applying differentiated weighted corrections, the deviations in occurrence projection for anomaly zones are effectively corrected. This ensures that occurrence information is consistent with the geological framework, improves the accuracy of orebody occurrence analysis, and provides reliable geological data support for rare earth mineral exploration and development.

[0077] In the specific analysis process: First, for each identified orebody occurrence anomaly segment, its occurrence anomaly characteristic information is extracted procedurally: 1. Morphological abrupt change degree: Calculate the normal distance between all boundary grid points of the segment and the trend surface fitted by the grid points of adjacent normal segments, and take the standard deviation of the distance as the abrupt change index. 2. Thickness change rate: Generate a vertical profile every 5 meters along the strike of the orebody, calculate the orebody thickness within the profile, and calculate the ratio of the thickness difference between adjacent profiles to the profile spacing, taking the maximum ratio as the thickness change rate of the segment. 3. Geophysical Response Instability: Extract the microseismic energy attenuation gradient and apparent resistivity values ​​of all grids within the section, calculate their relative deviations from the background average of the deposit, and take the average of the two relative deviations as the instability coefficient. Next, perform anomaly cause discrimination: Combine the above three characteristic values ​​into a feature vector and compare it with a preset cause-feature rule base. For example, the rule base defines: if "thickness change rate > 0.3 and morphological abrupt change index is high," and the section is located in a strong groundwater runoff zone in the hydrogeological information map, then it is judged as having a "hydrological activity influence" cause; if "geophysical response instability is high and morphological abrupt change index is moderate," and the section corresponds to a region with severe weathering crust base undulation in the ore body weathering crust information, then it is judged as having a "weathering crust interface change" cause. Then, perform differentiated weighted correction: based on the determined cause, call the preset weight coefficients (such as "hydrological activity"). The weights for "dynamic influence" (W_h=0.3), "weathering crust interface change" (W_w=0.4), and "tectonic interference" (W_t=0.6) are assigned. For each occurrence parameter (such as interface depth Z) that needs to be corrected within the anomaly segment, the correction amount ΔZ=(Z_observed-Z_inferred)*W, where Z_observed is the constraint value from the multi-source collaborative anomaly feature set (such as the depth corresponding to the extremely low resistivity point), and Z_inferred is the initial value in the ore body's three-dimensional morphology inference information. Finally, a consistency check is performed: the corrected ore body occurrence information is substituted into the basic geological spatial constraint framework, and it is checked whether the corrected ore body interface crosses known impermeable layers or stable rock layers, etc., hard constraint boundaries. If it crosses, the correction amount ΔZ is reduced proportionally until it does not cross, thus forming the final consistent ore body occurrence information.

[0078] In alternative or modified implementations: causal identification can be independent of a preset rule base, instead employing a pre-trained lightweight random forest classification model. The training samples for this model originate from anomaly segments with identified causes and their feature vectors in historical exploration data. The weight coefficient W in the differentiated weighted correction can be changed from a fixed value to a function related to the feature strength, such as W_t=0.4+0.3*sigmoid (morphological mutation index), allowing the weights to be dynamically adjusted according to the severity of the anomaly. In the coordination verification stage, in addition to hard boundary constraints, feedback from a multi-source collaborative anomaly feature set can be introduced: the theoretical geophysical response of the corrected model is calculated and compared with the actual observations, with fine-tuning iterations aimed at minimizing the root mean square error.

[0079] Figure 3 This is a schematic diagram of the structure of a rare earth underground orebody occurrence analysis system based on multidimensional sensor data provided in an embodiment of this application, as shown below. Figure 3 As shown, the rare earth underground ore body occurrence analysis system 300 based on multidimensional sensor data in this embodiment includes: a data alignment module 301, an inversion analysis module 302, and an adaptive analysis module 303.

[0080] The data alignment module 301 is used to acquire a multi-source heterogeneous sensor dataset. Based on the multi-source heterogeneous sensor dataset, it eliminates the deviations of different sensor data in sampling scale, timestamp, and spatial location to obtain spatiotemporally aligned multi-source sensor data. The inversion analysis module 302 is used to acquire a geological environment information set. Based on the geological environment information set and combined with the spatiotemporally aligned multi-source sensor data, it performs joint inversion analysis on the extension trend of ion-adsorption type rare earth ore bodies to obtain ore body extension trend inversion information. The adaptive analysis module 303 is used to perform adaptive analysis on abnormal sections of underground ore body occurrence based on the ore body extension trend inversion information, and generate and output a rare earth underground ore body occurrence analysis log containing three-dimensional results of ore body occurrence.

[0081] Optionally, when the data alignment module 301 eliminates the deviations in sampling scale, timestamps, and spatial locations of different sensor data based on the multi-source heterogeneous sensor dataset to obtain spatiotemporally aligned multi-source sensor data, it is specifically used for: the multi-source heterogeneous sensor dataset includes microseismic response data and electromagnetic detection data; based on the microseismic response data, discrete event timestamps are analyzed, and synchronization alignment is performed by introducing a unified time reference event to obtain time-unified microseismic data; based on the electromagnetic detection data, the coordinates of multi-source measuring points are analyzed, and scale normalization is performed through a unified grid spatial resampling rule to obtain spatial grid electromagnetic data; based on the time-unified microseismic data and combined with the spatial grid electromagnetic data, the spatial coordinates and temporal attributes of the two are registered and fused to obtain the spatiotemporally aligned multi-source sensor data.

[0082] Optionally, when the data alignment module 301 performs registration and fusion analysis on the spatial coordinates and temporal attributes of the time-unified microseismic data and the spatial grid electromagnetic data to obtain the spatiotemporally aligned multi-source sensing data, it is specifically used for: analyzing the spatial coordinate distribution of microseismic events based on the time-unified microseismic data, and mapping the coordinates of each microseismic event to the corresponding spatial grid cell based on the unified grid space followed by the spatial grid electromagnetic data to obtain a spatially registered microseismic event sequence; analyzing the unified time reference corresponding to each event based on the microseismic event sequence, and performing time window association matching on the microseismic events and electromagnetic detection results based on the detection time interval of the spatial grid electromagnetic data on the corresponding grid cell to obtain a set of spatiotemporally associated event pairs; and analyzing the correspondence between the microseismic event attributes and electromagnetic response characteristics within the same grid cell based on the set of spatiotemporally associated event pairs, and generating a feature description that fuses microseismic and electromagnetic information within each grid cell by superimposing and verifying the characteristics of both, thereby obtaining the spatiotemporally aligned multi-source sensing data.

[0083] Optionally, when the inversion analysis module 302 performs joint inversion analysis on the extension trend of ion-adsorption type rare earth ore bodies based on the geological environment information set and combined with the spatiotemporally aligned multi-source sensor data to obtain ore body extension trend inversion information, it is specifically used for: the geological environment information set including hydrogeological information and ore body weathering crust information; based on the hydrogeological information, analyzing the constraints of groundwater runoff on the ion migration and enrichment law of the ore body to obtain groundwater constraint information; based on the ore body weathering crust information, analyzing the constraints of weathering crust stratification on the ore body basement morphology and three-dimensional distribution framework to obtain weathering crust morphology information; integrating the groundwater... Water constraint information and weathering crust morphology information are used to construct a basic geological spatial constraint framework for the ore body. Based on the spatiotemporally aligned multi-source sensing data, the interaction between microseismic event attributes and electromagnetic response characteristics is analyzed. According to the synergistic change information of spatial attenuation of microseismic energy and anomalous response of electromagnetic apparent resistivity in the ore body occurrence area, a multi-source synergistic anomaly feature set is obtained. Based on the basic geological spatial constraint framework and combined with the multi-source synergistic anomaly feature set, the coupling relationship between geological constraints and geophysical anomalies in three-dimensional space is analyzed. The extension boundary and trend of the ore body are optimized, deduced, and delineated to obtain the ore body extension trend inversion information.

[0084] Optionally, the inversion analysis module 302, during the construction of the multi-source coordinated anomaly feature set, is specifically used for: analyzing the energy attenuation gradient characteristics of microseismic event attributes in each spatial grid cell during spatial propagation based on the time-unified microseismic data, to obtain microseismic energy attenuation information; analyzing the apparent resistivity change characteristics of electromagnetic response features in the same spatial grid cell in three-dimensional space based on the spatial grid electromagnetic data, to obtain electromagnetic apparent resistivity anomaly information; analyzing the synchronicity of changes of the two in the same spatial grid cell based on the microseismic energy attenuation information and the electromagnetic apparent resistivity anomaly information, and identifying coordinated anomaly zones related to ore body occurrence based on the regions where the increase in microseismic energy attenuation gradient and the decrease in electromagnetic apparent resistivity anomaly occur simultaneously in space; and analyzing the corresponding intensity and spatial continuity of the microseismic energy attenuation characteristics and electromagnetic apparent resistivity anomaly characteristics based on the coordinated anomaly zones, and obtaining the multi-source coordinated anomaly feature set through joint measurement and spatial aggregation of the two characteristics.

[0085] Optionally, when the inversion analysis module 302 analyzes the coupling relationship between geological constraints and geophysical anomalies in three-dimensional space, it is specifically used to: analyze the gradient change characteristics along the depth direction in each spatial grid cell based on the microseismic energy attenuation information to obtain microseismic gradient change information; analyze the anomalous fluctuation characteristics with increasing depth in the same spatial grid cell based on the electromagnetic apparent resistivity anomaly information to obtain electromagnetic anomaly fluctuation information; analyze the matching degree of the changing trends of the two in the corresponding spatial grid cell and depth interval based on the microseismic gradient change information and the electromagnetic anomaly fluctuation information, and identify synchronous change areas according to the degree of synergy between gradient increase and anomaly decrease in spatial location and change pattern; and analyze the continuity and correlation strength of microseismic gradient and electromagnetic anomaly fluctuation in three-dimensional space based on the synchronous change areas to obtain the coupling relationship.

[0086] Optionally, when the inversion analysis module 302 optimizes and delineates the extension boundary and trend of the ore body to obtain the inversion information of the ore body extension trend, it is specifically used for: analyzing the spatial stability and consistency information of the microseismic gradient change information and the electromagnetic anomaly fluctuation information based on the synchronous change area to obtain the mineralization clue area; analyzing the continuity and change trend of the morphology of the mineralization clue area in three-dimensional space according to the coupling relationship, and inferring the main extension direction of the ore body in the horizontal and vertical directions in combination with the basic geological spatial constraint framework to obtain the three-dimensional morphology deduction information of the ore body; and, based on the three-dimensional morphology deduction information of the ore body and the multi-source collaborative anomaly feature set, smoothing and optimizing the spatial contour of the ore body within the boundary of the basic geological spatial constraint framework to delineate the continuous distribution range of the ore body in three-dimensional space to obtain the inversion information of the ore body extension trend.

[0087] Optionally, when the adaptive analysis module 303 performs adaptive analysis on the abnormal sections of the underground ore body occurrence based on the ore body extension trend inversion information, and generates and outputs a rare earth underground ore body occurrence analysis log containing the three-dimensional results of the ore body occurrence, it is specifically used to: analyze the local abrupt changes and discontinuities in morphology, thickness, and geophysical response characteristics within the three-dimensional distribution range of the ore body based on the ore body extension trend inversion information, and obtain ore body occurrence abnormal section information; based on the ore body occurrence abnormal section information, combined with the basic geological spatial constraint framework, and according to the adaptive weighted correction of the abnormality cause, obtain ore body occurrence information; integrate the ore body extension trend inversion information, the ore body occurrence abnormal section information, and the corresponding ore body occurrence information to generate and output the rare earth underground ore body occurrence analysis log.

[0088] Optionally, the adaptive analysis module 303, during the construction of the ore body occurrence information, is specifically used for: analyzing the characteristics of each anomalous segment in terms of morphological abruptness, thickness variation rate, and geophysical response instability based on the ore body occurrence anomaly segment information, to obtain occurrence anomaly characteristic information; analyzing the correlation between different anomaly characteristics and geological environmental genesis based on the occurrence anomaly characteristic information, combined with the hydrogeological information and the ore body weathering crust information, to obtain anomaly cause discrimination information; performing differentiated weighted correction on the occurrence inference results of the anomalous segments based on the anomaly cause discrimination information, combined with the ore body three-dimensional morphology inference information, according to the influence weights of different genesis factors such as hydrological activity, weathering crust interface changes, and tectonic interference on the continuity of ore body occurrence, to obtain corrected ore body occurrence information; and analyzing the coordination between the correction results and the overall geological framework based on the corrected ore body occurrence information, combined with the basic geological spatial constraint framework, and obtaining the final coordinated ore body occurrence information through feedback verification with the multi-source collaborative anomaly feature set.

[0089] The system in this embodiment can be used to execute the methods of any of the above embodiments, and its implementation principle and technical effect are similar, so they will not be described again here.

Claims

1. A method for analyzing the occurrence of rare earth underground ore bodies based on multidimensional sensor data, characterized in that, include: A multi-source heterogeneous sensor dataset is acquired. Based on the multi-source heterogeneous sensor dataset, the biases of different sensor data in sampling scale, timestamp and spatial location are eliminated to obtain spatiotemporally aligned multi-source sensor data. A geological environment information set is acquired. Based on the geological environment information set and the spatiotemporally aligned multi-source sensing data, a joint inversion analysis is performed on the extension trend of ion adsorption type rare earth ore bodies to obtain ore body extension trend inversion information. Based on the inversion information of the ore body extension trend, an adaptive analysis is performed on the abnormal sections of the underground ore body occurrence, and a rare earth underground ore body occurrence analysis log containing the three-dimensional results of the ore body occurrence is generated and output.

2. The method according to claim 1, characterized in that, The process involves eliminating biases in sampling scale, timestamp, and spatial location among different sensor data based on the multi-source heterogeneous sensor dataset to obtain spatiotemporally aligned multi-source sensor data, including: The multi-source heterogeneous sensing dataset includes microseismic response data and electromagnetic detection data; Based on the microseismic response data, the timestamps of discrete events are analyzed, and time-uniform microseismic data is obtained by introducing a unified time reference event for synchronization and alignment. Based on the electromagnetic detection data, the coordinates of multi-source measuring points are analyzed, and scale normalization is performed through a unified grid space resampling rule to obtain spatial grid electromagnetic data. Based on the time-unified microseismic data and combined with the spatial grid electromagnetic data, the spatial coordinates and temporal attributes of the two are registered and fused to obtain the spatiotemporally aligned multi-source sensing data.

3. The method according to claim 2, characterized in that, The spatial coordinates and temporal attributes of the time-unified microseismic data, combined with the spatial grid electromagnetic data, are registered and fused to obtain the spatiotemporally aligned multi-source sensing data, including: Based on the time-unified microseismic data, the spatial coordinate distribution of microseismic events is analyzed. Combined with the unified grid space followed by the spatial grid electromagnetic data, the coordinates of each microseismic event are mapped to the corresponding spatial grid cell to obtain a spatially registered microseismic event sequence. Based on the microseismic event sequence, the unified time reference corresponding to each event is analyzed. Combined with the detection time interval of the spatial grid electromagnetic data on the corresponding grid cell, the microseismic events and electromagnetic detection results are correlated and matched with time windows to obtain a set of spatiotemporally correlated event pairs. Based on the set of spatiotemporally correlated event pairs, the correspondence between the attributes of microseismic events and the characteristics of electromagnetic response within the same grid cell is analyzed. By superimposing and verifying the consistency of the two characteristics, a feature description that integrates microseismic and electromagnetic information within each grid cell is generated, thus obtaining the spatiotemporally aligned multi-source sensing data.

4. The method according to claim 3, characterized in that, Based on the geological environment information set and combined with the spatiotemporally aligned multi-source sensor data, a joint inversion analysis is performed on the extension trend of ion-adsorption type rare earth ore bodies to obtain ore body extension trend inversion information, including: The geological environment information set includes hydrogeological information and ore body weathering crust information; Based on the aforementioned hydrogeological information, the constraints of groundwater runoff on the migration and enrichment patterns of ions in the ore body are analyzed to obtain groundwater constraint information. Based on the weathering crust information of the ore body, the constraints of weathering crust stratification on the ore body base morphology and three-dimensional distribution framework are analyzed to obtain weathering crust morphology information. By integrating the groundwater constraint information and the weathering crust morphology information, a basic geological spatial constraint framework for the ore body is constructed. Based on the spatiotemporally aligned multi-source sensing data, the interaction between microseismic event attributes and electromagnetic response characteristics is analyzed. According to the information on the coordinated changes of spatial attenuation of microseismic energy and abnormal response of electromagnetic apparent resistivity in the ore body occurrence area, a multi-source coordinated anomaly feature set is obtained. Based on the aforementioned basic geological spatial constraint framework, and combined with the multi-source collaborative anomaly feature set, the coupling relationship between geological constraints and geophysical anomalies in three-dimensional space is analyzed. The extension boundary and trend of the ore body are optimized, deduced, and delineated to obtain the inversion information of the ore body extension trend.

5. The method according to claim 4, characterized in that, The process of constructing the multi-source collaborative anomaly feature set includes: Based on the time-unified microseismic data, the energy attenuation gradient characteristics of microseismic event attributes in each spatial grid cell during spatial propagation are analyzed to obtain microseismic energy attenuation information. Based on the electromagnetic data of the spatial grid, the electromagnetic response characteristics within the same spatial grid cell are analyzed in terms of apparent resistivity variation characteristics in three-dimensional space to obtain electromagnetic apparent resistivity anomaly information. Based on the microseismic energy attenuation information and the electromagnetic apparent resistivity anomaly information, the synchronicity of their changes within the same spatial grid cell is analyzed. Based on the regions where the increase in microseismic energy attenuation gradient and the decrease in electromagnetic apparent resistivity anomaly occur simultaneously in space, the ore body-related synergistic anomaly regions are identified. Based on the aforementioned coordinated anomaly region, the corresponding intensity and spatial continuity of the microseismic energy attenuation characteristics and electromagnetic apparent resistivity anomaly characteristics are analyzed. By jointly measuring and spatially aggregating the characteristics of both, the multi-source coordinated anomaly feature set is obtained.

6. The method according to claim 5, characterized in that, The specific implementation of the analysis of the coupling relationship between geological constraints and geophysical anomalies in three-dimensional space includes: Based on the microseismic energy attenuation information, the gradient variation characteristics along the depth direction within each spatial grid cell are analyzed to obtain the microseismic gradient variation information. Based on the electromagnetic apparent resistivity anomaly information, the abnormal fluctuation characteristics with increasing depth within the same spatial grid cell are analyzed to obtain electromagnetic anomaly fluctuation information. Based on the microseismic gradient change information and the electromagnetic anomaly fluctuation information, the degree of matching between the two in the corresponding spatial grid cells and depth ranges is analyzed. Based on the degree of coordination between the gradient increase and the anomaly decrease in spatial location and change pattern, synchronous change areas are identified. Based on the synchronously changing region, the continuity and correlation strength of microseismic gradient and electromagnetic anomaly fluctuations in three-dimensional space are analyzed to obtain the coupling relationship.

7. The method according to claim 6, characterized in that, The optimization and delineation of the ore body's extension boundary and trend yields the ore body extension trend inversion information, including: Based on the synchronous change region, the spatial stability and consistency of the microseismic gradient change information and the electromagnetic anomaly fluctuation information are analyzed to obtain the mineralization clue region. Based on the coupling relationship, the continuity and change trend of the mineralization clue area in three-dimensional space are analyzed. Combined with the basic geological spatial constraint framework, the main extension direction of the ore body in the horizontal and vertical directions is inferred to obtain the three-dimensional morphology inference information of the ore body. Based on the three-dimensional morphological inference information of the ore body and the multi-source collaborative anomaly feature set, within the boundary of the basic geological spatial constraint framework, the spatial contour of the ore body is smoothly connected and the boundary is optimized to delineate the continuous distribution range of the ore body in three-dimensional space, thereby obtaining the ore body extension trend inversion information.

8. The method according to claim 7, characterized in that, The method involves adaptively analyzing anomalous sections of underground orebody occurrence based on the orebody extension trend inversion information, generating and outputting a rare earth underground orebody occurrence analysis log containing three-dimensional orebody occurrence results, including: Based on the inversion information of the ore body extension trend, the local abrupt changes and discontinuities in morphology, thickness and geophysical response characteristics within the three-dimensional distribution range of the ore body are analyzed to obtain information on the ore body occurrence anomaly sections. Based on the information on the abnormal occurrence of the ore body, combined with the basic geological spatial constraint framework, and according to the adaptive weighted correction of the cause of the anomaly, the occurrence information of the ore body is obtained. By integrating the ore body extension trend inversion information, the ore body occurrence anomaly section information, and the corresponding ore body occurrence information, the occurrence analysis log of the rare earth underground ore body is generated and output.

9. The method according to claim 8, characterized in that, The process of constructing the occurrence information of the ore body includes: Based on the information on the abnormal occurrence sections of the ore body, the characteristics of each abnormal section in terms of morphological abruptness, thickness change rate and geophysical response instability are analyzed to obtain the abnormal occurrence characteristic information. Based on the occurrence anomaly characteristics information, combined with the hydrogeological information and the weathering crust information of the ore body, the correlation between different anomaly characteristics and geological environmental causes is analyzed to obtain anomaly cause discrimination information. Based on the anomaly cause discrimination information and the three-dimensional morphology inference information of the ore body, the occurrence inference results of the anomaly section are differentiated and weighted according to the influence weights of different causes of hydrological activity, weathering crust interface changes and tectonic interference on the continuity of ore body occurrence, so as to obtain the corrected ore body occurrence information. Based on the corrected ore body occurrence information and combined with the basic geological spatial constraint framework, the consistency between the correction result and the overall geological framework is analyzed. Through feedback verification with the multi-source collaborative anomaly feature set, the final consistent ore body occurrence information is obtained.

10. A rare earth underground orebody occurrence analysis system based on multidimensional sensor data, characterized in that, The method applied to any one of claims 1-9 includes: The data alignment module is used to acquire multi-source heterogeneous sensor datasets and, based on the multi-source heterogeneous sensor datasets, eliminate the deviations of different sensor data in sampling scale, timestamp and spatial location to obtain spatiotemporally aligned multi-source sensor data. The inversion analysis module is used to acquire a geological environment information set. Based on the geological environment information set and combined with the spatiotemporally aligned multi-source sensor data, a joint inversion analysis is performed on the extension trend of the ion adsorption type rare earth ore body to obtain the ore body extension trend inversion information. The adaptive analysis module is used to perform adaptive analysis on abnormal sections of the underground ore body occurrence based on the inversion information of the ore body extension trend, and generate and output a rare earth underground ore body occurrence analysis log containing three-dimensional results of the ore body occurrence.

Images