Search

CN-122022059-A - Knowledge-graph-based ore-forming prediction system

CN122022059ACN 122022059 ACN122022059 ACN 122022059ACN-122022059-A

Abstract

The invention provides an ore-forming prediction system based on a knowledge graph, and belongs to the technical field of mines. The mining area disaster risk analysis system comprises a data acquisition module, a knowledge graph construction module, a risk assessment module and a risk response module, wherein the data acquisition module is used for constructing a mining area multisource monitoring data set, the knowledge graph construction module is used for generating a mining area disaster knowledge graph for representing the association relation between the underground structure state of a mining area and disaster induction factors, the risk knowledge graph construction module is used for constructing a disaster risk knowledge graph for representing the hidden geological disaster characteristics of the mining area, the risk assessment module is used for constructing collapse risk coefficients, constructing stress concentration coefficients and inducing earthquake sensitivity coefficients, the risk response module is used for constructing comprehensive risk coefficients, presetting comprehensive risk thresholds, comparing the comprehensive risk coefficients with the comprehensive risk thresholds, triggering alarm instructions and optimizing the comprehensive risk coefficients. Features such as mining area activities are reflected from multiple dimensions, so that integral characterization of mining area hidden geological disaster risks is realized, judgment is not performed by means of a single monitoring index, and mining safety of the mining area is improved.

Inventors

  • YANG CHEN
  • JIAO SEN
  • Cheng Hongzhu
  • LI YINZHEN
  • WANG HUANZHI

Assignees

  • 中化地质矿山总局山东地质勘查院

Dates

Publication Date
20260512
Application Date
20260210

Claims (10)

  1. 1. An ore-forming prediction system based on a knowledge graph is characterized by comprising: The data acquisition module is used for acquiring a rock stratum resistivity value, a rock stratum density real-time measurement value, a roadway surrounding rock displacement, a microseismic energy release value and a rock mass strain gauge measurement value corresponding to the drilling depth of the mining area under the application scene of monitoring the geological disasters of the mining area, so as to construct a mining area multisource monitoring data set; The knowledge graph construction module is used for carrying out entity modeling, relationship association and semantic mapping by utilizing the mining area multisource monitoring data set to generate a mining area disaster knowledge graph representing the association relationship between the mining area underground structure state and disaster induction factors; The risk knowledge image construction module is used for carrying out aggregation analysis on the structural state entity corresponding to the mining area and the association relation thereof based on the mining area disaster knowledge map to construct a disaster risk knowledge image for describing the hidden geological disaster characteristics of the mining area; The risk assessment module is used for carrying out fusion analysis on the surrounding rock stability state, the electrical abnormal state, the density abnormal state, the strain concentration state and the microseismic energy evolution state formed by mapping the rock stratum resistivity value, the rock stratum density real-time measurement value, the roadway surrounding rock displacement, the microseismic energy release value and the rock mass strain gauge measurement value based on the disaster risk knowledge image to construct a collapse risk coefficient Structural stress concentration coefficient And an induced seismic sensitivity coefficient ; Risk response module, which is to collapse risk factor Structural stress concentration coefficient And an induced seismic sensitivity coefficient Combining to construct a comprehensive risk coefficient Presetting a comprehensive risk threshold SC, and integrating a risk coefficient In contrast to the comprehensive risk threshold SC, when And when SC, triggering an alarm instruction and optimizing.
  2. 2. The knowledge-graph-based ore formation prediction system according to claim 1, wherein the data acquisition module comprises a rock stratum resistivity value acquisition unit, a rock stratum density real-time measurement value acquisition unit, a roadway surrounding rock displacement amount acquisition unit, a microseismic energy release value acquisition unit and a rock mass strain gauge measurement value acquisition unit; The system comprises a rock stratum resistivity value acquisition unit, a resistivity measurement result and a drilling depth one by one to form a continuous profile curve, a standard resistance module is used for carrying out zero point and measuring range calibration on the logging probe before acquisition, and meanwhile, marking and correcting abnormal jump points in the acquisition process to obtain the rock stratum resistivity value corresponding to the drilling depth; The real-time measurement value acquisition unit is used for lowering a density logging probe into a borehole in the drilling construction process or after the drilling is completed, measuring the attenuation characteristic of the rock stratum to rays by a gamma ray emission and scattering receiving mode, and converting the attenuation characteristic into a volume density value in real time according to the corresponding relation between the attenuation intensity and the stratum density, compensating and correcting the well diameter change, the well fluid density and the environmental interference factors in the measuring process, synchronously recording a density measurement result and the borehole depth to form a continuous density curve, and verifying the data stability by repeated measurement to obtain the real-time measurement value of the rock stratum density.
  3. 3. The knowledge graph-based ore prediction system is characterized in that the roadway surrounding rock displacement acquisition unit is used for arranging surrounding rock displacement monitoring points in a roadway target section of a mining area, installing convergence measurement reference points on two sides of a roadway section and a top and bottom plate, periodically measuring distance changes among the reference points through a laser range finder or a section convergence meter to acquire surrounding rock convergence displacement, simultaneously drilling monitoring holes in the surrounding rock and embedding a multipoint displacement meter to continuously monitor the relative displacement of surrounding rocks with different depths, carrying out zero point correction and abnormal drift correction on displacement data in the acquisition process, and binding and storing the displacement, roadway mileage position and time stamp so as to acquire the roadway surrounding rock displacement; The micro-seismic energy release value acquisition unit is used for arranging a plurality of three-component geophones or acceleration sensors in the range of a mining area, realizing multi-point synchronous sampling through a unified time service system, continuously monitoring the rock mass vibration signal of the mining area, automatically identifying a micro-seismic event and intercepting corresponding waveform data when the vibration signal meets a triggering condition, denoising and filtering the waveform, calculating an energy integral value in an effective time window or obtaining an event energy release value according to the conversion of the magnitude of vibration, and combining a plurality of station data to perform fusion correction so as to improve the calculation accuracy, thereby obtaining the micro-seismic energy release value; the rock mass strain gauge measurement value acquisition unit is used for drilling a monitoring hole in a stress key position of surrounding rock of a mining area and burying the rock mass strain gauge, so that the strain gauge and the rock mass form a tight coupling structure, after grouting fixation and initial zero calibration are completed, a data acquisition device is used for reading a strain gauge output signal according to a set period, meanwhile, temperature compensation parameters are acquired for correcting a measurement result, and corrected strain data are stored in association with a spatial position and a time stamp, so that the rock mass strain gauge measurement value is obtained, and a mining area multi-source monitoring data set is constructed.
  4. 4. The knowledge-based ore-forming prediction system according to claim 3, wherein the knowledge-graph construction module comprises a modeling unit, a relationship association unit and a semantic mapping unit; The modeling unit is used for performing entity modeling by using the mining area multisource monitoring data set; Firstly, analyzing a rock stratum resistivity value, a rock stratum density real-time measurement value, a roadway surrounding rock displacement, a microseismic energy release value and a rock mass strain gauge measurement value in multi-source monitoring data set of a mining area according to a monitoring source, a collecting position and a monitoring object type, and aggregating data records with the same monitoring object identification and space attribution characteristics; On the basis, based on a preset entity type dividing rule, mapping the aggregated data records into a drilling entity, a roadway entity, a monitoring point entity, a rock stratum entity, a microseismic event entity and a surrounding rock entity respectively, defining a corresponding attribute set for each type of entity, and carrying physical parameter information related to the entity in the mining area multisource monitoring data set, and forming a structured expression facing a mining area object by the data in the mining area multisource monitoring data set in a logical mapping mode by attaching entity type identification, entity unique identification and attribute field mapping relation to the data records so as to establish a mining area entity-attribute model; the relation association unit is used for constructing association relations among different entity types in the mining area multisource monitoring data set based on the mining area entity-attribute model; Firstly, matching entity nodes positioned in the same space region based on space position attributes corresponding to the entity nodes, and constructing a space association relation between the entities; secondly, associating the entity nodes with time sequence relation based on the time attribute corresponding to the entity nodes, and constructing a time evolution association relation between the entities; On the basis, based on the change consistency among the physical parameters of different entity nodes, the entity nodes are associated, and a parameter coupling association relation among the entities is constructed; the semantic mapping unit is used for carrying out semantic mapping on the rock stratum resistivity value, the rock stratum density real-time measurement value, the roadway surrounding rock displacement, the microseismic energy release value and the rock mass strain gauge measurement value in the mining area multisource monitoring data set based on the mining area entity-attribute model; Firstly, dividing intervals of rock stratum resistivity values, rock stratum density real-time measurement values, roadway surrounding rock displacement amounts, microseismic energy release values and rock mass strain gauge measurement values according to preset parameter semantic mapping rules, and mapping continuous numerical parameters into surrounding rock stability states, electrical abnormal states, density abnormal states, strain concentration states and microseismic energy evolution states; secondly, attaching the stability state, the electrical abnormal state, the density abnormal state, the strain concentration state and the microseismic energy evolution state of the surrounding rock to corresponding entity nodes as semantic attributes for representing the state characteristics of the entity nodes at the current monitoring moment; and constructing a mining area disaster knowledge graph of the association relationship between the underground structure state of the mining area and disaster inducing factors by fusing the entity nodes, the association relationship among the entities and semantic attribute information.
  5. 5. The knowledge-graph-based ore-forming prediction system of claim 4, wherein the risk knowledge representation construction module comprises a structural state entity screening unit, an association relation aggregation unit and a risk representation generation unit; The structure state entity screening unit is used for screening entity nodes bearing surrounding rock stability states, electrical abnormal states, density abnormal states, strain concentration states and microseismic energy evolution states based on the mining area disaster knowledge graph; The association relation aggregation unit is used for merging the structural state entities with the same spatial attribution based on the spatial association relation, the time evolution association relation and the parameter coupling association relation between the structural state entities, and combining and summarizing the corresponding structural state semantics; And the risk image generation unit is used for expressing the aggregated structural state semantics in a multidimensional characteristic form to form a disaster risk knowledge image for integrally describing the hidden geological disaster characteristics of the mining area.
  6. 6. The knowledge-graph-based ore-forming prediction system of claim 5, wherein the risk assessment module comprises a fusion analysis unit, a collapse risk calculation unit, a collapse risk assessment unit, a stress concentration calculation unit, a stress concentration assessment unit, and an evoked seismic sensitivity calculation unit and an evoked seismic sensitivity assessment unit; The fusion analysis unit is used for performing time alignment and space alignment treatment on the stability state, the electrical abnormal state, the density abnormal state, the strain concentration state and the microseismic energy evolution state of the surrounding rock, and the states are comparable in a unified space unit and a time window; the collapse risk calculation unit is used for calculating the rock stratum resistivity value R, the roadway surrounding rock displacement D and the rock mass strain gauge measurement value based on the mining area multisource monitoring data set The collapse risk factor is obtained by ; Firstly, calculating surrounding rock deformation strength term for rock stratum resistivity value R ; Calculating surrounding rock deformation strength item by using roadway surrounding rock displacement D ; Using rock mass strain gauge measurements Calculating strain accumulation term ; Then based on surrounding rock deformation strength item And a strain accumulation term Computing nonlinear coupling enhancement terms ; Finally based on surrounding rock deformation strength item Deformation strength term of surrounding rock Strain accumulation term Nonlinear coupling enhancement In combination, the collapse risk factor is calculated by ; ; In the formula, And Respectively denoted as weight coefficients.
  7. 7. The knowledge-graph-based ore-forming prediction system according to claim 6, wherein the collapse risk assessment unit is configured to preset a collapse risk threshold AX and combine the collapse risk threshold AX with a collapse risk coefficient Comparison is made, comprising: When (when) When AX is greater than AX, the stability of surrounding rock is abnormal in the mining area at the current monitoring time, the operation load or the propulsion rate is required to be adjusted downwards by 10-30%, the supporting parameters of the surrounding rock are adjusted, the supporting rigidity is improved by 15-40%, and the data acquisition frequency of key monitoring points is increased by 20-50%; When (when) And when the AX is less than or equal to the AX, the surrounding rock stability of the mining area is normal at the current monitoring moment, a collapse risk early warning instruction is not triggered, and the current mining parameters and the supporting strategy are maintained.
  8. 8. The knowledge-graph-based ore-formation prediction system of claim 7, wherein the stress concentration calculation unit is configured to measure formation density in real time based on multi-source monitoring data set of the mining area Rock mass strain gauge measurements And roadway surrounding rock displacement D, a structural stress concentration coefficient is obtained by the following way ; First, real-time formation density measurements are used Calculating density outliers ; Using rock mass strain gauge measurements Computing strain concentration gradient terms ; Then calculating a displacement non-uniformity item by using the roadway surrounding rock displacement D ; Next, the density anomaly term is added And strain concentration gradient term In combination, the stress coupling term is calculated by the following formula ; Finally, the density anomaly item Gradient term for strain concentration Non-uniform displacement term Stress coupling term The structural stress concentration coefficient is calculated by the following formula ; ; The stress concentration evaluation unit is used for presetting a structural stress concentration threshold value AD and concentrating the structural stress concentration coefficient In contrast to the stress concentration threshold AD, comprising: When (when) When AD is carried out, the mining area is abnormal in structural stress concentration at the current monitoring moment, the mining operation intensity is reduced by 10-30%, the supporting rigidity is adjusted to be improved by 15-40%, and the continuous operation time interval is prolonged by 15-35%; When (when) And when the stress concentration of the mining area is less than or equal to AD, the stress concentration of the mining area is normal, a stress slow-release optimization strategy is not executed, and the mining area continues to operate according to the established mining operation and supporting scheme.
  9. 9. The knowledge-based ore prediction system according to claim 8, wherein the evoked seismic sensitivity calculation unit is configured to calculate the microseismic energy release value E and the rock strain gauge measurements based on the mining area multisource monitoring dataset And surrounding rock deformation strength term And with density anomaly terms In combination, the evoked seismic sensitivity coefficients are obtained by ; Firstly, calculating a micro-seismic energy active term by utilizing a micro-seismic energy release value E ; ; Using rock mass strain gauge measurements Calculating strain accumulation rate term ; ; Subsequently, the density anomaly term is utilized Computing structural non-homogeneous magnification term ; ; Finally, the surrounding rock deformation strength term Micro-vibration energy active term Rate of strain accumulation term And a structural inhomogeneity magnification term Fusion is carried out, and the induced seismic sensitivity coefficient is obtained through the calculation of the following formula ; ; The induced earthquake sensitivity evaluation unit is used for presetting an induced earthquake sensitivity threshold YC and carrying out induced earthquake sensitivity coefficient In contrast to the evoked seismic sensitivity threshold YC, comprising: When (when) When YC, the mining area is abnormal in mining disturbance under the current monitoring time, the mining area mining operation propulsion speed is required to be adjusted downwards by 20% -40%, the data acquisition frequency of microseismic monitoring is improved by 30% -60%, and the mining operation time window is adjusted, so that the continuous operation interval time is prolonged by 15% -35%; When (when) And when the YC is less than or equal to the YC, the response of the mining area to the mining disturbance at the current monitoring moment is in a stable state, the disturbance inhibition and the monitoring strengthening strategy are not executed, and the mining area continues to operate according to the reference operation parameters and the monitoring frequency.
  10. 10. The knowledge-graph-based ore-forming prediction system of claim 9, wherein the risk response module comprises an association unit, a comprehensive calculation unit, and an optimization unit; the associated unit is used for carrying out collapse risk factor Structural stress concentration coefficient And an induced seismic sensitivity coefficient Combining and carrying out normalization treatment to obtain normalized collapse risk coefficients respectively Structural stress concentration coefficient And an induced seismic sensitivity coefficient ; The comprehensive calculation unit is used for calculating the collapse risk coefficient based on Structural stress concentration coefficient And an induced seismic sensitivity coefficient The comprehensive risk coefficient is obtained through calculation according to the following formula ; ; The optimizing unit is used for integrating the comprehensive risk threshold value SC with the comprehensive risk coefficient Comparison is made, comprising: When (when) When SC, the underground structure state of the mining area is abnormal at the current monitoring time, an alarm instruction is triggered, an optimization strategy is executed, the data acquisition frequency of key monitoring points of the mining area is increased by 20% -50%, the intensity of the mining operation in progress in the mining area is adjusted downwards, the operation load or the advancing rate is reduced by 10% -30%, and surrounding rock supporting parameters are adjusted aiming at a risk area, so that the supporting rigidity is increased by 15% -40%; When (when) And when SC is less than or equal to the SC, the underground structure state of the mining area under the current monitoring is normal, an alarm instruction is not triggered, and the current monitoring configuration and operation control strategy are maintained.

Description

Knowledge-graph-based ore-forming prediction system Technical Field The invention relates to the technical field of mines, in particular to an ore-forming prediction system based on a knowledge graph. Background With the increasing exhaustion of mineral resources, mining activities of mining areas gradually develop to deep and complex geological environments, and in the mining process of the deep mining areas, due to complex and changeable geological conditions, geological disaster risks of different degrees, particularly problems of collapse, stress concentration, earthquake induction and the like are often accompanied, so that the disaster risks of the mining areas are effectively predicted and evaluated, the safety and stability of the mining area mining processes are ensured, and the mining area geological disaster monitoring technology is developed. Traditional mining area geological disaster monitoring is generally based on single monitoring data sources, such as geological radar, strain gauges, displacement meters and other single-point monitoring means, the systems can only provide local information, the data processing mode is more traditional, data islanding phenomenon exists, fusion analysis of different data sources cannot be effectively realized, global disaster assessment capability is lacked, accuracy and timeliness of disaster early warning are poor, and safety management of mining areas is affected. Disclosure of Invention In order to overcome the above disadvantages, the present invention provides an ore-forming prediction system based on a knowledge graph, which overcomes the above technical problems or at least partially solves the above problems. The invention is realized in the following way: the invention provides an ore-forming prediction system based on a knowledge graph, which comprises the following steps: The data acquisition module is used for acquiring a rock stratum resistivity value, a rock stratum density real-time measurement value, a roadway surrounding rock displacement, a microseismic energy release value and a rock mass strain gauge measurement value corresponding to the drilling depth of the mining area under the application scene of monitoring the geological disasters of the mining area, so as to construct a mining area multisource monitoring data set; The knowledge graph construction module is used for carrying out entity modeling, relationship association and semantic mapping by utilizing the mining area multisource monitoring data set to generate a mining area disaster knowledge graph representing the association relationship between the mining area underground structure state and disaster induction factors; The risk knowledge image construction module is used for carrying out aggregation analysis on the structural state entity corresponding to the mining area and the association relation thereof based on the mining area disaster knowledge map to construct a disaster risk knowledge image for describing the hidden geological disaster characteristics of the mining area; The risk assessment module is used for carrying out fusion analysis on the surrounding rock stability state, the electrical abnormal state, the density abnormal state, the strain concentration state and the microseismic energy evolution state formed by mapping the rock stratum resistivity value, the rock stratum density real-time measurement value, the roadway surrounding rock displacement, the microseismic energy release value and the rock mass strain gauge measurement value based on the disaster risk knowledge image to construct a collapse risk coefficient Structural stress concentration coefficientAnd an induced seismic sensitivity coefficient; Risk response module, which is to collapse risk factorStructural stress concentration coefficientAnd an induced seismic sensitivity coefficientCombining to construct a comprehensive risk coefficientPresetting a comprehensive risk threshold SC, and integrating a risk coefficientIn contrast to the comprehensive risk threshold SC, whenAnd when SC, triggering an alarm instruction and optimizing. In a preferred scheme, the data acquisition module comprises a rock stratum resistivity value acquisition unit, a rock stratum density real-time measurement value acquisition unit, a roadway surrounding rock displacement acquisition unit, a microseismic energy release value acquisition unit and a rock mass strain gauge measurement value acquisition unit; The system comprises a rock stratum resistivity value acquisition unit, a resistivity measurement result and a drilling depth one by one to form a continuous profile curve, a standard resistance module is used for carrying out zero point and measuring range calibration on the logging probe before acquisition, and meanwhile, marking and correcting abnormal jump points in the acquisition process to obtain the rock stratum resistivity value corresponding to the drilling depth; The real-time measurement value acquisition uni