CN-122022609-A - Modeling method based on digital twin mine

CN122022609ACN 122022609 ACN122022609 ACN 122022609ACN-122022609-A

Abstract

The invention relates to the technical field of digital twinning, in particular to a modeling method based on a digital twinning mine, which comprises the following steps of S1, collecting geological survey data, equipment operation data and material transportation data of the mine, and S2, carrying out space-time alignment and noise elimination on the geological survey data, the equipment operation data and the material transportation data to form a comprehensive data set. According to the invention, through carrying out space-time alignment, noise rejection and interpolation complement on multi-source data of mine geological survey, equipment operation and material transportation, the problem of messy and inconsistent traditional modeling data is solved, and a high-quality data base is provided for model construction. The causal relationship matrix is generated through the causal inference algorithm, so that the dynamic causal relationship among geological change, equipment operation and transportation scheduling is accurately depicted, and the defect that the traditional model cannot accurately reflect the real-time cooperative relationship of the three is overcome.

Inventors

ZHANG JIANQIANG
NING SHUZHENG
ZHANG DEGAO
XUE HUANHUAN
NIU FANGPING
HOU YUEHUA
ZHOU YAO

Assignees

中国煤炭地质总局勘查研究总院

Dates

Publication Date: 20260512
Application Date: 20251230

Claims (10)

1. A digital twin mine based modeling method, comprising: s1, collecting geological survey data, equipment operation data and material transportation data of a mine; s2, carrying out space-time alignment and noise elimination on the geological survey data, the equipment operation data and the material transportation data to form a comprehensive data set; S3, constructing a mine three-dimensional geological dynamic model, an equipment intelligent adaptation model and a transportation dynamic collaborative model based on the comprehensive data set, and generating a causal relationship matrix representing dynamic causal relationship among geological change, equipment operation and transportation scheduling through a causal inference algorithm; s4, generating a model calibration rule according to the causal relationship matrix and the current mining scene, and primarily updating an intelligent adaptation model and a transportation dynamic collaborative model of the equipment; S5, performing virtual simulation previewing on the preliminary updating scheme, identifying potential conflicts and generating a conflict resolution scheme; S6, performing incremental updating on the three-dimensional geological dynamic model of the mine, the intelligent adaptation model of the equipment and the transportation dynamic collaborative model according to the conflict resolution scheme, and establishing a full-link traceability file; and S7, performing precision evaluation on the updated model through real-time feedback data of the physical mine, and starting a closed-loop evolution process when the evaluation result does not reach the standard.
2. The digital twin mine modeling method according to claim 1, wherein the S2 specifically comprises: Performing time sequence alignment on the multi-source data by using a time stamp synchronization protocol; Unifying the data under different coordinate systems to a mine global coordinate system by adopting a space coordinate conversion algorithm; identifying and eliminating abnormal data points by applying a sliding window filter or an isolated forest algorithm; And interpolating and complementing the missing data based on the geological relevance of the adjacent areas, the running state trend of the equipment and the transportation path circulation logic to form the comprehensive data set.
3. The digital twin mine modeling method according to claim 1, wherein in S3, a causal relationship matrix characterizing dynamic causal relationships among geological changes, equipment operation and transportation scheduling is generated by a causal inference algorithm, and specifically comprises: defining geologic state variables Device state variables Transport state variable Observations at time step t; Calculating the causal strength between variables by adopting a Grangel causal test or a structural equation model; Constructing a causal correlation matrix Wherein the elements are A causal influence coefficient indicating the i-th state to the j-th state, i, j e { G, M, T }; Setting causal conduction threshold τ when When, it is confirmed that there is an active causal link from the i-th to the j-th state.
4. The digital twin mine modeling method according to claim 1, wherein in S4, generating a model calibration rule according to the causal relationship matrix and the current mining scenario specifically includes: a historical scene database containing different mining stages, geological conditions and production loads is constructed in advance; Extracting scene categories with similarity to the current real-time geological state, the equipment operation data and the transportation load data meeting a preset threshold value from the historical scene database; Invoking a calibration parameter set corresponding to the scene category from a preset calibration rule library, wherein the calibration parameter set comprises a model parameter calibration threshold, a causal conduction efficiency coefficient and a cooperative response time threshold; and applying the calibration parameter set to the causal relation matrix to generate a model calibration rule applicable to the current scene.
5. The digital twin mine modeling method according to claim 1, wherein the step S5 specifically comprises: Loading equipment performance limit parameters and a transportation network topological structure by taking the state of the current three-dimensional geological dynamic model of the mine as an initial condition; Advancing in a time sequence in a simulation environment, and performing simulation on the preliminary updating scheme; monitoring whether surrounding rock stress overrun, equipment load rate overrun or transportation path congestion index exceeds a preset safety threshold value in the simulation process; If any overrun condition is monitored, judging that potential conflict exists, starting a multi-objective optimization solver, and generating at least two feasible conflict resolution schemes by taking mining efficiency, a safety threshold, cost control and equipment service life as optimization targets.
6. The digital twin mine modeling method according to claim 5, wherein the starting the multi-objective optimization solver, with the mining efficiency, the safety threshold, the cost control and the equipment life as optimization objectives, generates a plurality of possible conflict resolution schemes, and specifically comprises: defining a multi-objective optimization function: ; wherein x is the decision variable vector to be optimized, Representing loss of production efficiency, safety risk index, operating cost increment and equipment loss rate, Is a corresponding target weight vector; solving a pareto front solution set under the constraint condition g (x) is less than or equal to 0, wherein the constraint condition comprises equipment physical limit, transport network capacity and geological safety boundary; And screening solutions meeting causal link matrix verification conditions from the pareto front solution set to serve as the conflict resolution scheme.
7. The digital twin mine modeling method according to claim 1, wherein the step S6 specifically comprises: identifying model nodes related to changes in the conflict resolution scheme and causal association influence ranges thereof; only parameter adjustment or structure modification is carried out on the model nodes and the affected associated nodes, and incremental update is completed; And generating a unique identifier for the updating operation, and recording information of the type of an updating trigger source, the version number of a causal relation matrix, a matched calibration rule identifier, input and output data of simulation previewing, decision variable values of a conflict resolution scheme and a finally adopted updating instruction sequence to a traceability file.
8. The modeling method based on digital twin mines according to claim 1, wherein in S7, the accuracy of the updated model is evaluated by real-time feedback data of the physical mines, specifically comprising: Continuously acquiring actual geological deformation data, real-time working condition data of equipment and position data of a transport vehicle through an intelligent sensing system deployed in a physical mine; comparing the output data of the actual acquired data with the output data of the virtual model under the same space-time coordinates; Calculating causal relationship matching degree Accuracy of data mapping Decision execution bias rate Three evaluation indexes; when any evaluation index exceeds a preset tolerance range, the evaluation accuracy evaluation does not reach the standard.
9. The digital twin mine modeling method of claim 8, wherein the calculating causal relationship matching degree Accuracy of data mapping Decision execution bias rate The three evaluation indexes specifically comprise: Calculating causal relationship matching degree The method comprises the following steps: ; Wherein N is the number of sampling points, As the actual state change amount of the physical entity at the kth sampling point, Predicting state change quantity for the virtual model based on the causal relation matrix; Calculating the data mapping accuracy The ratio of the number of the effective matching data points to the total number of the data points is; calculating a decision execution bias rate And normalizing the Euclidean distance between the actual execution action and the recommended action of the model.
10. The modeling method based on digital twin mines according to claim 1, wherein in S7, a closed loop evolution process is started when the evaluation result does not reach the standard, specifically comprising: positioning a deviation root link based on the causal correlation matrix to reversely track a deviation propagation path, wherein the deviation root link comprises at least one of data acquisition, causal modeling, calibration rule matching or simulation previewing; aiming at the deviation root link, performing targeted optimization, namely adjusting the sampling frequency of sensing equipment or calibrating sensor parameters if the data acquisition problem is solved, retraining a causal inference model and updating a causal association matrix if the causal modeling problem is solved, and supplementing a new scene sample in a historical scene database and reconstructing a calibration rule base if the causal inference model is solved; And taking the optimized data, model or rule as new input, re-executing the complete process from the integrated data set construction to the model increment update, and completing one closed loop evolution iteration.

Description

Modeling method based on digital twin mine Technical Field The invention relates to the technical field of digital twinning, in particular to a modeling method based on a digital twinning mine. Background The digital twin technology becomes a core support for digital transformation of mines, and the visual control and optimization of the whole elements and the whole process of the mines are realized by constructing a real-time interactive channel of a virtual space and a physical mine. The technology can integrate multi-source data such as mine geology, equipment, transportation and the like to form a dynamically updated virtual model, provides data support for mining planning, safety control and the like, and is one of key technologies for intelligent mine and green mine construction. The prior disclosed invention patent CN115047833A provides a mine digital twin factory and a construction method thereof, and provides a mine intelligent sensing system, a digital twin model and a multi-time space scale database, and intelligent management and control of a mine are realized through virtual-real interaction. The patent mainly solves the problems of fusion and integration of multi-source heterogeneous data, establishes a complete twin architecture, and does not pay attention to the collaborative mapping precision of different dimension data in the dynamic model updating process. In the prior art, mine modeling needs to integrate dynamic data of multiple dimensions such as ore body structure, equipment operation, material transportation and the like, and model layers corresponding to data of different dimensions are often constructed based on respective data dimensions. When the model is dynamically updated, the real-time change of the ore body structure deviates from the mapping of the equipment running state and the material transportation path, the real-time cooperative relationship among the three cannot be accurately reflected, the accuracy of exploitation decision and the rationality of production scheduling are further affected, the problem is particularly obvious in complex mining scenes of mines, and no targeted solution thought is given by the existing scheme. Disclosure of Invention Aiming at the defects of the prior art, the invention provides a modeling method based on a digital twin mine, which solves the problems of the prior art such as the collaborative mapping deviation of multidimensional data, the insufficient suitability of dynamic model updating and the complete link tracing missing. In order to achieve the purpose, the invention is realized by the following technical scheme that the modeling method based on the digital twin mine comprises the following steps: s1, collecting geological survey data, equipment operation data and material transportation data of a mine; s2, carrying out space-time alignment and noise elimination on the geological survey data, the equipment operation data and the material transportation data to form a comprehensive data set; S3, constructing a mine three-dimensional geological dynamic model, an equipment intelligent adaptation model and a transportation dynamic collaborative model based on the comprehensive data set, and generating a causal relationship matrix representing dynamic causal relationship among geological change, equipment operation and transportation scheduling through a causal inference algorithm; s4, generating a model calibration rule according to the causal relationship matrix and the current mining scene, and primarily updating an intelligent adaptation model and a transportation dynamic collaborative model of the equipment; S5, performing virtual simulation previewing on the preliminary updating scheme, identifying potential conflicts and generating a conflict resolution scheme; S6, performing incremental updating on the three-dimensional geological dynamic model of the mine, the intelligent adaptation model of the equipment and the transportation dynamic collaborative model according to the conflict resolution scheme, and establishing a full-link traceability file; and S7, performing precision evaluation on the updated model through real-time feedback data of the physical mine, and starting a closed-loop evolution process when the evaluation result does not reach the standard. Further, the step S2 specifically includes: Performing time sequence alignment on the multi-source data by using a time stamp synchronization protocol; Unifying the data under different coordinate systems to a mine global coordinate system by adopting a space coordinate conversion algorithm; identifying and eliminating abnormal data points by applying a sliding window filter or an isolated forest algorithm; And interpolating and complementing the missing data based on the geological relevance of the adjacent areas, the running state trend of the equipment and the transportation path circulation logic to form the comprehensive data set. Further, in the step S3, a causal