A method for adaptive design of blasting parameters and collaborative execution with drilling rigs based on multi-source data

Abstract

This invention provides a method for adaptive design of blasting parameters and collaborative execution of drilling rigs based on multi-source data, relating to the field of intelligent mining technology. This method introduces multi-source raw measurement data, such as measurement while drilling, geophysical exploration, geological logging, and 3D scanning, as system inputs to construct an automated design mechanism for blasting parameters and a dynamic correction mechanism for construction deviations. It also utilizes post-blast measurement feedback to self-learn and update the design model, thereby achieving quantifiable, verifiable, and iteratively optimized blasting processes. This invention does not rely on traditional human experience and judgment, proactively avoiding the influence of subjective human factors on the formulation of blasting parameters and construction effects, ensuring the scientific nature, consistency, and repeatability of parameter design.

Description

Technical Field

[0001] This invention relates to the field of intelligent mining technology, specifically to a method for adaptive design of blasting parameters and collaborative execution of drilling rigs based on multi-source data. Background Technology

[0002] For hard rock metal mines, drill-and-blast remains the most important and economical method for rock breaking and mining. However, the traditional drill-and-blast operation mode has long suffered from systemic shortcomings, failing to meet the core requirements of intelligent and refined transformation and upgrading of mines. This directly slows down the overall digital and intelligent development of mines and has become a core obstacle to improving the quality and efficiency of drill-and-blast operations.

[0003] From the perspective of the detailed stages of drilling and blasting operations, there are prominent pain points in the three core stages of design, construction, and post-blasting evaluation, making it difficult to achieve refined and standardized operations. The blasting parameter design stage relies entirely on experience, with a severe lack of scientific quantification. Parameter formulation depends on the personal experience of technical personnel and static design manuals. The coupling mechanism between geology and blasting is unclear, and there is a lack of refined blastability inversion models supported by multi-source data. Key blasting parameters lack quantitative mapping relationships with geological conditions and blasting targets, and excellent operational experience cannot be structurally accumulated and reused. The drilling construction stage suffers from uncontrolled deviations and a lack of closed-loop correction. The "testing-drilling" link is severely disconnected, and parameters such as hole location, depth, and inclination cannot be monitored and fed back in real time. Data flow in drilling and blasting collaboration is poor, key data during construction is not effectively collected, and potential drilling quality problems directly translate into blasting quality issues. The post-blasting evaluation stage relies mainly on manual experience judgment, with a complete lack of quantitative feedback. Core evaluation indicators are scored manually by visual inspection, which is highly subjective. The correlation between post-blasting effects and deviations in the initial design and construction cannot be traced, and no self-learning optimization mechanism has been formed. Repeated trial and error for similar problems is very common. The root cause of the pain points in the entire drilling and blasting operation process mentioned above lies in the dual absence of a multi-source heterogeneous data fusion mechanism and an intelligent decision-making model, which in turn forms a systemic technical bottleneck throughout the drilling and blasting operation. The invention patent with publication number CN120541368A provides a multi-source data fusion preprocessing method for blasting design. However, this method only achieves the fusion and preprocessing of multi-source data, without establishing a quantitative correlation model between the fused data and the blasting parameter design, and lacks a feedback correction mechanism based on post-blasting effects. It cannot achieve closed-loop optimization between blasting design, construction execution, and effect evaluation, resulting in a certain degree of reliance on experience and insufficient adaptability in blasting parameter design.

[0004] In view of this, it is urgent to break through the bottlenecks in key technologies such as data fusion, intelligent design, drilling and blasting coordination, and quantitative evaluation of effects in the field of drilling and blasting in hard rock metal mines. It is necessary to study an integrated drilling and blasting technology method adapted to intelligent mines to completely solve the technical problems of traditional operation mode. Summary of the Invention

[0005] In view of the technical problems existing in the background art, the present invention provides a method for adaptive design of blasting parameters and collaborative execution of drilling rig based on multi-source data. By deeply integrating multi-source data such as drilling measurement, geophysical exploration, geological information and 3D scanning, a breakability inversion model, an automated blasting parameter design model and a feedback self-learning model are constructed to realize the automated generation, fine matching and collaborative execution of blasting parameters with intelligent drilling rig.

[0006] In a first aspect, embodiments of the present invention provide a method for adaptive design of blasting parameters and collaborative execution of drilling rigs based on multi-source data, which includes the following steps:

[0007] S1, Data Acquisition and Input: Acquire multi-source data of the blasting area, including drilling measurement data, geophysical data, 3D scanning data, mining site geological data, and historical construction record data;

[0008] S2, Data Fusion Analysis: The multi-source data are spatiotemporally fused to generate a fused feature vector; based on the fused feature vector, an integrated learning model is used to invert the rock mass mechanical parameters and the breakability index.

[0009] S3, Blasting parameter optimization design: Based on the lithological parameters and breakability index, a blasting design scheme including at least hole mesh parameters and charge parameters is generated through a multi-objective optimization algorithm;

[0010] S4, Drilling and blasting coordinated execution: The blasting parameter set is automatically mapped into drilling rig control commands and charge initiation and detonation coordinated commands and issued for execution; Real-time collection of construction measurement data, identification of design-construction deviations and dynamic correction of the charge initiation and detonation coordinated commands;

[0011] S5, Blasting effect feedback update: Obtain post-blast measured effect data, quantify the deviation between blasting design parameters and actual blasting effect, and use the deviation information as feedback input to introduce data fusion analysis and blasting parameter optimization design process, update relevant model parameters, thereby providing a basis for subsequent blasting parameter design and forming a closed-loop optimization.

[0012] As a further improvement of the present invention, in step S2, the data fusion analysis adopts a multi-source fusion algorithm based on adaptive weight control, specifically including:

[0013] S21, Spatiotemporal Synchronization Model Construction: Construct a unified four-dimensional spatiotemporal index ID=f(x,y,z,t), and establish a nearest neighbor matching structure for multi-source data using a kd-tree structure combined with timestamp sorting. This structure is applied to each data point D in the multi-source data. jSpatial distance and time difference are calculated, and spatiotemporal weights are calculated based on a Gaussian weighting function to achieve spatiotemporal synchronization of multi-source data; wherein, the calculation formula for the spatiotemporal weights is as follows:

[0014] ;

[0015] Among them, TS j For spatiotemporal weights; D j For the data points participating in the matching; d s d represents spatial distance; t For time difference; and These are the spatial and temporal scale parameters, respectively; C j The initial confidence level for candidate data sources;

[0016] S22, Feature Mapping: Constructing Feature Value Matching F for Candidate Data Points j An embedding transformation method is used to map discrete categories into computationally computable numerical features. After embedding, all categorical variables form a uniform vector space, which, together with continuous variables, constitutes a multi-source feature vector, ultimately yielding a fused feature vector F. sync The formula is as follows:

[0017] ;

[0018] S23, Confidence Management: Based on the weight distribution of each data source participating in the fusion and the variance of candidate data points, the variance σ of the synchronization result is calculated through unbiased estimation. sync A confidence management mechanism is constructed; wherein, the variance σ sync The calculation formula is as follows:

[0019] ;

[0020] S24, Anomaly Suppression: When the variance of the synchronization result exceeds the threshold, the number of data points participating in the fusion is insufficient, or the weight concentration is too low, the anomaly suppression mechanism is triggered to mark the synchronization anomaly unit and perform compensation processing.

[0021] As a further improvement of the present invention, in step S2, the specific process of lithological classification and ruptureability inversion in the data fusion analysis is as follows:

[0022] S25 employs a combination of an ensemble learning model and an optimized classification tree method, inputting the cleaned multi-source feature vectors into the ensemble learning model for inverting the lithological strength index K. v Fragility index K p The integrated learning model, while completing the lithological parameter inversion, further constructs a blasting effect prediction model to establish the mapping relationship between the blasting parameter set Θ and the block size distribution, vibration response, and cross-section forming results.

[0023] S251, the ensemble learning backbone model adopts XGBoost, GBDT or LightGBM gradient boosting ensemble model, and constructs multiple weak learning trees through boosting iteration to achieve high-precision classification of lithology and continuous prediction of breakability index.

[0024] S252, optimize the classification tree auxiliary model as the structural constraint and assistant of the backbone model. Select CART classification tree or cost-sensitive classification tree to perform structured hierarchical discrimination of the response of key lithology categories in blasting.

[0025] S253, a weight control and feature fusion mechanism, introduces a confidence-based weight control and multi-layer feature fusion mechanism during training and inference. The model assigns weight factors that are updated in real time to each feature based on its contribution to the explosion effect prediction. This achieves adaptive adjustment to strengthen high-contribution features and weaken low-contribution features; among which, the weighting factor The calculation formula is as follows:

[0026] ;

[0027] Among them, g i The influence of each feature value obtained from gradient boosting calculation on the prediction of blasting effect.

[0028] As a further improvement of the present invention, step S3, the optimization design of the blasting parameters specifically includes:

[0029] S31, using the fused feature vector F sync and lithological strength index K v Fragility index K p As input, a multi-objective optimization function is used, and the output includes hole spacing S, row spacing B, hole depth H, charge amount Q, and charge structure correction coefficient q. （z） and extended network The optimization function for the multi-objective parameter space of the multiple sets of blasting parameters is as follows:

[0030] ;

[0031] Among them, J frag The block size index model predicts the difference between the block size and the target block size in the project; J energy As an indicator of energy utilization rationality, the constraint function avoids unreasonable parameter combinations that result in high energy consumption and low efficiency; E( E represents the input energy generated per unit of rock breaking corresponding to the current parameter combination. d (k) v k p The required energy is obtained by inversion based on lithological parameters and breakability indices. , Let be the weight coefficients of the objective function, and satisfy . Used to balance block size control and energy utilization efficiency; The weighting coefficient for the i-th spatial unit or sampling region is determined based on the volume ratio or ore grade of that region. This is a set of blasting parameters.

[0032] As a further improvement of the present invention, step S3, the optimization design of the blasting parameters, further includes:

[0033] S32 introduces safety and forming constraints:

[0034] ;

[0035] Among them, J vib The vibration velocity of particles at different distances from the blast center is used to constrain blasting parameters to meet safety control requirements; J shape An index representing the predicted deviation between the profile and the cross-sectional shape after blasting. To allow for relative deviations between over- and under-excavation, used to constrain over- and under-excavation trends and guide the optimization of forming quality; A pred The predicted cross-sectional profile and area after blasting; A design This represents the design values ​​for the cross-sectional profile and area after blasting; v pred This is the predicted value of particle vibration; v limit These are the safety limits for particle vibration.

[0036] S33, after obtaining multiple sets of candidate blasting parameter schemes that meet the optimization conditions, based on the output results of the blasting effect prediction model described in step S25, a comprehensive executability scoring function Score for the blasting scheme is further constructed. The scoring function is defined as:

[0037] Score = λ1·Frag Score +λ2·Energy Score +λ3·Vib Score +λ4·Shape Score ;

[0038] Among them, Frag Score This is an evaluation value constructed based on the deviation between the predicted block size distribution and the target block size according to the current blasting scheme. The evaluation value is calculated based on the average block diameter, the rate of oversized blocks, or the difference in the block size distribution curve, and is used to reflect the effectiveness of block size control. (Energy) Score This is an evaluation value constructed based on the degree of matching between the unit rock-breaking energy consumption of the current blasting scheme and the theoretical optimal energy consumption. The evaluation value is based on the actual input energy and the theoretical energy consumption E. d The deviation is calculated and used to measure energy utilization efficiency; VibScore This is an evaluation value constructed based on the relationship between predicted vibration velocity and safety threshold according to the current blasting scheme. The evaluation value is based on the predicted particle vibration velocity v. pred With the allowable vibration velocity v limit The deviation is calculated and used to assess blasting safety; Shape Score The evaluation value is constructed based on the deviation between the post-blast cross-sectional profile and area of ​​the current blasting scheme and the designed cross-section. This evaluation value is calculated according to the cross-sectional area deviation rate or profile fitting error and is used to reflect the forming quality; λ 1、 λ 2、 λ 3、 λ4 represents the weighting coefficients for each evaluation value, satisfying: It is used to achieve a comprehensive optimization goal that balances block size control, energy utilization, safety, and molding quality;

[0039] Each individual score is obtained by normalizing, classifying, or mapping the prediction results, and is used to transform the blasting effect index into an engineering evaluation value under a unified dimension.

[0040] As a further improvement of the present invention, in step S4, the drilling and blasting coordinated execution specifically includes:

[0041] S41, Drilling execution command generation: Based on the blasting parameter set Θ={S,B,H,Q,q} (z) ,Dn}, automatically generate drilling execution instructions for single hole level and hole network level; the instructions include at least the drilling start and end coordinates, drilling inclination angle and azimuth angle and allowable attitude deviation range, single hole design depth and over-depth compensation amount, drilling speed and propulsion force and rotation speed control range, and drilling type identifier.

[0042] S42, Generation of Charge and Initiation Coordination Commands: Based on the optimization results of blasting parameters and lithology inversion results, further generate charge and initiation coordination commands, including single-hole charge amount Q and charge structure segmentation parameters q. (z) , charge density and coupling state and cavity structure control parameters, initiation method selection, delay network Dn and inter-hole row delay time configuration;

[0043] S43, Collect and transmit construction status data: Through the drilling rig control system, drilling measurement system and blasting control terminal, collect and transmit construction status data in real time, including actual borehole trajectory and depth, borehole direction and attitude change information, propulsion force and rotation speed, drilling pressure and torque and footage speed during drilling, abnormal signals in the borehole, actual charge quantity and charge structure execution status and detonation parameter configuration status; compare and analyze the actual construction data with the original design command parameters, automatically identify borehole geometric deviation, drilling process deviation, charge execution deviation, and detonation network deviation, and output quantitative deviation amplitude, duration and spatial distribution characteristics.

[0044] S44, Automatic Design-Construction Deviation Identification: Compares and analyzes construction status data with original design command parameters to automatically identify deviations, including borehole geometric deviations, drilling process deviations, charge execution deviations, and detonation network deviations.

[0045] As a further improvement of the present invention, in step S5, the blasting effect feedback update is used to quantitatively evaluate the actual effect after the blasting operation is completed, and the evaluation result is used as a parameter feedback signal input to the data fusion analysis and blasting parameter optimization to realize the self-learning update and closed-loop optimization of the blasting parameter design model, specifically including:

[0046] S51. After the blasting is completed, collect the measured data of the blasting effect, including blasting vibration monitoring data, post-blast block size distribution data, blasting profile data, and post-blast energy utilization information.

[0047] S52, after registering the collected data with a unified timestamp and spatial coordinates, a set of measured features of the blasting effect is formed, and a unified blasting deviation feature vector is constructed, as follows:

[0048] ;

[0049] ;

[0050] Among them, e frag For block size deviation index; e shape For contour deviation index; e vib For blasting vibration deviation index; e energy As an indicator of blasting energy utilization rate, J is used as the reference in step S2. energy The feedback correction amount of the energy rationality constraint function is used to update the empirical coefficients or correction terms of the unit rock-breaking energy consumption model E(Θ); D pred To predict the average block size for the model; D real The average size of the fragments was measured after the explosion; A pred To predict the post-explosion cross-sectional area for the model; A real v represents the measured cross-sectional area after the explosion. pred Predict peak vibration velocity for the model; v real E represents the measured peak vibration velocity after the blast. in E is the total energy input for the blast. eff For effective rock-breaking energy;

[0051] S53, the blasting deviation feature vector is introduced as a feedback input into the lithology inversion stage to characterize the degree of deviation between the design parameter output and the actual blasting effect. This vector is then introduced into the S2 lithology and breakability inversion algorithm stage to update the prediction weights of each parameter index. This provides support for the next step of inversion.

[0052] As a further improvement of the present invention, in the feature mapping described in step S22, One-Hot encoding or Embedding conversion is used to map discrete categories into digital features that can participate in the calculation; wherein, One-Hot encoding is suitable for variables with a small number of categories and clear mutual exclusion relationships; Embedding conversion embedding vectors are suitable for situations with a large number of categories, similarity between categories, or the need to express continuous implicit semantics.

[0053] As a further improvement of the present invention, step S24, the specific process of anomaly suppression, is as follows: if any of the following conditions occur, anomaly identification is triggered and marked as a "synchronization anomaly unit":

[0054] Case 1, synchronous variance σ sync Greater than the set threshold σ thr ;

[0055] Scenario 2, Number N of data points participating in the fusion <N min ;

[0056] Case 3, the weight concentration is too low (Max( ) / Sum( )< );

[0057] Case 4: Spatial distance or time difference exceeds the set tolerance range, d s >R tol or d t >T tol This could lead to potential cross-hole or cross-segment errors;

[0058] Furthermore, synchronization control is performed on the synchronization anomaly units, including: reducing confidence weight, triggering compensation retrieval, extending the time buffer, and calling interpolation predictions of similar data, to ensure the availability of the final synchronization results and avoid interference from mismatches of different spatiotemporal data on subsequent lithology inversion and blasting parameter design.

[0059] Secondly, embodiments of the present invention provide a system for adaptive design of blasting parameters and collaborative execution of drilling rigs based on multi-source data, which is used to execute the above-mentioned method for adaptive design of blasting parameters and collaborative execution of drilling rigs based on multi-source data, including:

[0060] The data acquisition module is used to acquire multi-source data from the blasting area, including at least drilling measurement data.

[0061] The fusion analysis module is used to perform spatiotemporal synchronous fusion of the multi-source data to generate a fusion feature vector; based on the fusion feature vector, an integrated learning model is used to invert the rock mass mechanical parameters and the breakability index.

[0062] The blasting parameter optimization design module is used to generate a blasting design scheme that includes at least hole mesh parameters and charge parameters based on the rock mass mechanics parameters and through a multi-objective optimization algorithm.

[0063] The drilling and blasting collaborative execution module is used to automatically map the blasting design scheme into executable instructions for the drilling rig and send them to the drilling rig control system to execute the drilling operation;

[0064] The blasting feedback update module is used to acquire post-blast measured effect data, construct a deviation feature vector, and update the model weights used in the fusion analysis module and the blasting parameter design module based on the deviation feature vector to form a closed-loop optimization.

[0065] As a further improvement of the present invention, the fusion analysis module includes:

[0066] The spatiotemporal synchronization unit is used to construct a four-dimensional spatiotemporal index structure and calculate the spatiotemporal weight of each data point based on a Gaussian weighting function.

[0067] The feature mapping unit is used to construct feature value matching for candidate data points, mapping discrete categories into numerical features that can participate in computation, and obtaining a fused feature vector.

[0068] The confidence management unit is used to calculate the variance of the synchronization results through unbiased estimation based on the initial confidence and data point variance of each data source.

[0069] The anomaly suppression unit is used to trigger anomaly marking and compensation processing when the variance of the synchronization result exceeds a threshold, the number of data points participating in the fusion is insufficient, or the weight concentration is too low.

[0070] Beneficial effects:

[0071] At the technical level, this invention achieves a fundamental shift from experience-based to data-driven approaches. By constructing a high-precision fusion and intelligent inversion mechanism for multi-source data, it significantly improves the accuracy of geological parameter acquisition and the scientific nature of blasting parameter design. At the same time, with the help of automatic design-execution mapping and real-time correction of construction deviations, it ensures the precise controllability of the entire process and the continuous evolution of the model, forming a closed-loop self-optimization capability.

[0072] From an economic perspective, this invention effectively reduces the consumption of blasting materials through energy utilization optimization and precise design of the charge structure, significantly shortens the operation cycle through automation of drilling operations and improved construction efficiency, and reduces subsequent processing costs through over- and under-excavation control. Thus, it achieves substantial savings in comprehensive costs in blasting, excavation, transportation and support, bringing considerable economic benefits.

[0073] At the safety and social level, this invention effectively controls hazards such as blasting vibration and flyrock through delayed network optimization and multi-objective constraints, reducing the risk of surrounding rock disturbance and collapse. At the same time, it reduces noise, dust and carbon emissions generated by blasting, significantly improving construction safety and environmental friendliness. In addition, it transforms implicit experience into explicit data models, promoting the standardization and intelligent transformation of blasting operations, reducing reliance on the experience of senior personnel, and providing strong support for the sustainable development of the industry.

[0074] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, and in order to make the above and other objects, features and advantages of the present invention more apparent and understandable, specific embodiments of the present invention are described below. Attached Figure Description

[0075] To more clearly illustrate the technical solution of the present invention, the accompanying drawings used in the present invention will be briefly described below. Obviously, the drawings described below are merely some embodiments of the present invention, and those skilled in the art can obtain other drawings based on these drawings without any creative effort.

[0076] Figure 1 This is a flowchart illustrating the adaptive design of blasting parameters and the collaborative execution of drilling rigs based on multi-source data, provided in an embodiment of the present invention. Detailed Implementation

[0077] The embodiments of the technical solution of the present invention will now be described in detail with reference to the accompanying drawings. These embodiments are merely illustrative of the technical solution of the present invention and are therefore intended to limit the scope of protection of the present invention.

[0078] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains; the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the invention; the terms “comprising” and “having”, and any variations thereof, in the specification, claims, and foregoing description of the invention, are intended to cover non-exclusive inclusion.

[0079] In the description of the embodiments of this invention, technical terms such as "first" and "second" are used only to distinguish different objects and should not be construed as indicating or implying relative importance or implicitly specifying the number, specific order, or primary and secondary relationship of the indicated technical features. In the description of the embodiments of this invention, "multiple" means two or more, unless otherwise explicitly defined.

[0080] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of the invention. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0081] In the description of the embodiments of this invention, the term "and / or" 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, and B existing alone. Additionally, the character " / " in this document generally indicates that the preceding and following related objects have an "or" relationship.

[0082] In the description of the embodiments of the present invention, the term "multiple" refers to two or more (including two), similarly, "multiple groups" refers to two or more (including two groups), and "multiple pieces" refers to two or more (including two pieces).

[0083] In the description of the embodiments of the present invention, the technical terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," "counterclockwise," "axial," "radial," and "circumferential" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing the embodiments of the present invention and simplifying the description, and are not intended to indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the embodiments of the present invention.

[0084] In the description of the embodiments of the present invention, unless otherwise explicitly specified and limited, the technical terms such as "installation," "connection," "joining," and "fixing" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. Those skilled in the art can understand the specific meaning of the above terms in the embodiments of the present invention according to the specific circumstances.

[0085] To address the technical challenge of traditional drilling and blasting methods failing to meet the demands of modern mines' development towards intelligence, precision, visualization, and automation, and the inability to establish a closed-loop optimization system centered on "measured data-driven, automated design, precise execution, and effect feedback," this invention provides a method for adaptive design of blasting parameters and collaborative execution by drilling rigs based on multi-source data fusion. This closed-loop optimization system for drilling and blasting parameters, based on "measured data-driven, automated design, precise execution, and effect feedback," introduces multi-source raw measurement data, including measurement while drilling, geophysical exploration, geological logging, and 3D scanning, as system inputs. It constructs an automated design mechanism for blasting parameters and a dynamic correction mechanism for construction deviations, and utilizes post-blast measured feedback to self-learn and update the design model, thereby achieving quantifiable, verifiable, and iterative optimization of the entire blasting process. This invention does not rely on traditional human experience and judgment, proactively avoiding the influence of subjective human factors on the formulation of blasting parameters and construction effects, ensuring the scientific nature, consistency, and repeatability of parameter design.

[0086] This invention targets the mining face of underground metal mines and is applicable to mining blasting operations in sublevel open-pit mining, shallow-hole ore-holding, or backfilling mining methods, but is not limited to any specific mining method.

[0087] Example 1

[0088] Please refer to Figure 1 As shown, Embodiment 1 of this invention discloses a method for adaptive design of blasting parameters and collaborative execution of drilling rigs based on multi-source data fusion. This embodiment selects a segmented backfilling mining method stope in an underground metal mine as the application object. The stope is buried at a depth of approximately 300–450 m, and the surrounding rock of the working face is generally medium to hard rock, with local development of weak interlayers and structural fracture zones. The integrity and breakability of the surrounding rock vary significantly, and the stability of the blasting effect is required to be high. The method mainly includes five working steps: data acquisition and input, data fusion analysis, blasting parameter optimization design, output of drilling and blasting collaborative execution instructions (construction process monitoring and deviation identification), and collection and feedback of blasting effect.

[0089] S1, before the blasting design, performs data acquisition steps, mainly including:

[0090] 1) Basic equipment parameters: These include the drilling rig coordinate system, drill rod length, drill arm kinematic parameters, drilling attitude range, and drilling rig communication protocol type; explosive type, charge density, detonation energy, detonation velocity, and initiation method. This embodiment mainly collects relevant parameters of the currently used rock drilling equipment, including the coordinate system parameters of the rock drilling rig, drill rod length, drill arm kinematic parameters, allowable drilling attitude range, and communication protocol type; it also records basic blasting information such as the explosive type, charge density, detonation velocity, and initiation method, which serve as the basic constraints for calculating blasting parameters and generating execution commands.

[0091] 2) Geological information of the stope: including geological logging data, core test parameters (uniaxial compressive strength, elastic modulus, joint density, shear strength), ore body occurrence conditions (ore body location, thickness, grade), surrounding rock grade, structural surface development, and average wave velocity of the stope; This embodiment combines the geological logging data of the stope with the results of core tests to obtain the magnitude of the stress in the stope, the mechanical parameters such as the uniaxial compressive strength, elastic modulus, shear strength, and average wave velocity of the ore body and surrounding rock, and to clarify the occurrence location, thickness, grade of ore, and grade distribution of the surrounding rock, providing basic input data for lithological identification and breakability inversion.

[0092] 3) Geophysical and monitoring data: including but not limited to distribution of electrical anomalies, transient electromagnetic characteristics, microseismic activity, and vibration source location distribution, used to provide prior information on deep geological structures and potential alternating hard and weak zones; this embodiment introduces existing geophysical and monitoring data from the mining area, including distribution of electrical anomalies, transient electromagnetic response characteristics, and microseismic monitoring results; by analyzing the spatial location of anomalies and vibration source distribution characteristics, potential faults, fracture zones, and weak interlayer areas are identified, providing prior information support for multi-source data fusion and identification of geological conditions ahead.

[0093] 4) Measurement while drilling (MWD) data, including torque, thrust, drill pressure, rotational speed, footage rate, energy consumption per unit footage, in-hole vibration signal, and drill pipe attitude time sequence data.

[0094] 5) Three-dimensional laser scanning data: used to acquire a point cloud model of the working face and extract structural features such as concavity / convexity, joint orientation, borehole location, and spatial geometric errors. This embodiment mainly performs three-dimensional laser scanning on the target working area to acquire a high-precision point cloud model of the working face and assigns a unified spatial coordinate reference to the point cloud data; based on the point cloud features, it extracts the joint and fracture development characteristics, structural surface orientation, and local hydrological anomaly indication information of the working face, providing geometric constraints for predicting blasting forming quality and borehole placement.

[0095] 6) Construction record data, including historical borehole mesh design, charge structure, delay network configuration, blasting effect records, etc.

[0096] After data collection is complete, the data is entered into the PC interactive interface in the form of an Excel spreadsheet and a project bin file. The data entered in this case is shown in the table below.

[0097]

[0098] S2, Data Fusion Analysis: This mainly includes multi-source data fusion algorithms, lithology and breakability inversion algorithms, and blasting parameter optimization calculation and prediction algorithms. After completing the multi-source basic data acquisition before the operation, data from different sources are uniformly input into the PC interactive interface to enter the multi-source data fusion analysis module. The system first establishes a unified spatiotemporal index framework, using borehole spatial coordinates, borehole number, and drilling depth as the core index dimensions, and automatically matches and synchronously associates 3D laser scanning point cloud data, geological logging and rock mechanics parameters, geophysical and microseismic monitoring information, and MWD (Measuring While Drilling) data.

[0099] To address the issues of inconsistent dimensions, sampling frequencies, and data precision among data from different sources, this invention employs a multi-source fusion algorithm based on adaptive weight adjustment. Its core structure includes: a spatiotemporal synchronization model, a feature mapping module, a confidence management module, and an anomaly suppression module.

[0100] S21, Spatiotemporal Synchronization Model Construction (Spatiotemporal Alignment and Quality Control of Multi-Source Data): To ensure the real-time adaptability of multi-source data during drilling, the system constructs a unified four-dimensional spatiotemporal index: ID=f(x,y,z,t). A nearest neighbor matching structure for multi-source data is established using a kd-tree and timestamp sorting, enabling automatic retrieval of geological parameters corresponding to drilling depth and spatiotemporal consistency control between real-time and historical data during the construction process. Specifically, this includes matching data points D... j Calculate spatial distance d s With time difference d t The spatiotemporal weights TS are obtained based on the Gaussian weighting function. j :

[0101] ;

[0102] in, and These are the spatial and temporal scale parameters, C. j The initial confidence level for candidate data sources (related to sensor accuracy or data quality).

[0103] In this embodiment, a matching data point is selected with a spatial distance ds = 0.5 m, a time difference dt = 2 s, a spatial scale parameter σs = 1.0 m, a time scale parameter σt = 5 s, and an initial confidence score Cj = 0.8 for the candidate data source. Substituting these values ​​into the formula, we obtain TSj ≈ 0.651, indicating that this data point has a high weight in multi-source fusion and can participate in subsequent feature construction.

[0104] S22, Feature mapping, constructing feature value matching F for candidate data points. jThis value is the formatted preprocessed physical features including, but not limited to, geological information, wave velocity, fault indication value, MWD single-point mechanical parameters, geophysical inversion data, point cloud location information, geometric errors, etc.

[0105] When processing input data of different categories (including lithology, surrounding rock grade, structural surface type, pore type, etc.), it is preferable to use One-Hot encoding or Embedding transformation to map discrete categories into numerical features that can participate in the calculation. Among them, One-Hot encoding is suitable for variables with a small number of categories and clear mutual exclusion relationships; Embedding transformation embedding vectors are suitable for situations with a large number of categories, similarity between categories, or the need to express continuous implicit semantics.

[0106] After embedding, all categorical variables form a uniform vector space, which, together with continuous variables, constitutes a structurally complete multi-source feature vector. The final effective feature representation after synchronous fusion is F. sync :

[0107] ;

[0108] In this embodiment, Fsync = all categorical variables, after embedding, form a uniform vector space, which, together with the continuous variables, constitutes a structurally complete multi-source feature vector. The final effective feature representation after synchronous fusion is: Fsync = [f1, f2, ..., fn]. In this embodiment, the fused feature vector after normalization is:

[0109] Fsync = [

[0110] 1, 0, 0,

[0111] 0, 1, 0, 0,

[0112] 0.72,

[0113] 0.68,

[0114] 0.75,

[0115] 0.40,

[0116] 0.65,

[0117] 0.55,

[0118] 0.60,

[0119] 0.70,

[0120] 0.45,

[0121] 0.50,

[0122] 0.60,

[0123] 0.58,

[0124] 0.35, 0.12 ]

[0127] The values ​​of the aforementioned continuous characteristic variables are processed from the original engineering parameters (such as uniaxial compressive strength, elastic modulus, wave velocity, etc.) using the minimum-maximum normalization method, mapping the original dimensional data to the [0,1] interval.

[0128] S23, Confidence management, based on the weight distribution of the data points involved in the fusion and the variance σ of the candidate data points. j The variance σ of the synchronization results is calculated using unbiased estimation. sync :

[0129] ;

[0130] In this embodiment, three data points are selected for fusion, with spatiotemporal weights of TS1=0.6, TS2=0.3, and TS3=0.1, respectively. The squared deviations between the corresponding feature values ​​and the mean of the fused features are as follows:

[0131] ;

[0132] The synchronous variance is then calculated as follows:

[0133] ;

[0134] S24, Abnormal Suppression: If any of the following conditions occur, abnormal identification is triggered and the unit is marked as a "synchronization abnormal unit":

[0135] Synchronous variance σ sync Greater than the set threshold σ thr ;

[0136] Number of data points N participating in the fusion <N min ;

[0137] The weight concentration is too low (Max( ) / Sum( )< );

[0138] Spatial distance or time difference exceeding the set tolerance range (d) s >R tol or d t >T tol This can lead to potential cross-hole or cross-segment errors.

[0139] Synchronization control is performed on abnormal units, including: reducing confidence weight, triggering compensation retrieval, extending time buffer, and calling interpolation predictions of similar data, to ensure the availability of the final synchronization results and avoid interference from mismatches of different spatiotemporal data on subsequent lithology inversion and blasting parameter design.

[0140] In this embodiment, Max(TS) j =0.6, ΣTS j If the value is 1.0, the weight concentration is 0.6, which is greater than the threshold γ=0.4, and no anomaly is triggered.

[0141] The above processing ensures that the fusion results are within a controllable variance range, thereby guaranteeing the consistency of data from different sources at the borehole scale or spatial unit scale.

[0142] After data alignment is completed, the system constructs multi-source fused feature vectors according to borehole depth segments or spatial grid cells. These feature vectors comprehensively characterize the mechanical properties, structural development state, and drilling response characteristics of the rock mass, and their contents include, but are not limited to:

[0143] 1. Rock mass mechanical characteristic parameters (uniaxial compressive strength, elastic modulus, shear strength, wave velocity, etc.);

[0144] 2. Structural and geometric characteristic parameters (joint and fracture density, structural plane orientation, magnitude of site stress, etc.);

[0145] 3. Geophysical exploration and monitoring response characteristics (electrical anomaly intensity, transient electromagnetic response characteristics, frequency and energy index of microseismic activity);

[0146] 4. Drilling process response characteristics (MWD parameters such as thrust, drill pressure, torque, drilling speed and energy consumption per unit footage).

[0147] To avoid interference from differences in feature scales or reliability with the inversion results, the system introduces a weight constraint mechanism for the fused features. Based on the stability, real-time performance, and historical verification results of the data source, corresponding weight coefficients are assigned to various features to achieve adaptive adjustment of feature contribution.

[0148] S25, the lithology and breakability inversion algorithm, after obtaining the fused feature vector, the system executes lithology identification and breakability index inversion processes. Lithology identification automatically distinguishes the lithology category of different depth sections or spatial units of the borehole through a classification model based on fused features and historical labeled samples. The breakability index inversion comprehensively reflects the rock mass strength level, structural integrity, and drilling response characteristics, and is used to characterize the ease of rock mass fracturing under blasting conditions. The breakability index is output in continuous or hierarchical form, and its value corresponds to the difficulty of rock mass fracturing, providing a quantitative basis for the subsequent zoned differentiated design of blasting parameters. While completing the lithology parameter inversion, the ensemble learning model further constructs a blasting effect prediction model to establish a blasting parameter set. The mapping relationship between block size distribution, vibration response and cross-sectional forming results.

[0149] The cleaned multi-source feature vectors are input into the supervised learning algorithm model for lithology K inversion. v With the breakability index K p .

[0150] Preferably, the lithology classification and breakability inversion algorithm of the present invention adopts the "combination method of ensemble learning model and optimized classification tree". The classification inversion model consists of two parts: an ensemble learning backbone model and an optimized tree structure auxiliary model.

[0151] S251 is an ensemble learning backbone model, preferably using gradient boosting ensemble models such as XGBoost, GBDT, or LightGBM. It constructs multiple weak learning trees through boosting iteration to achieve high-precision classification of lithology categories and continuous prediction of breakability index.

[0152] S252, Optimize the classification tree auxiliary model as a structural constraint and assistant of the backbone model. This invention can use CART classification tree or cost-sensitive classification tree to perform structured hierarchical discrimination of the response of key lithology categories such as "fracture zone, weak zone, and hard rock zone" in blasting.

[0153] S253, Weight Control and Feature Fusion Mechanism: To fully leverage the synergistic effect of the backbone ensemble model and the auxiliary classification tree, this invention introduces a confidence-based weight control and multi-layer feature fusion mechanism during training and inference. The model assigns real-time updated weight factors to each feature based on its contribution to the explosion effect prediction. This achieves an adaptive adjustment that strengthens high-contribution features and weakens low-contribution features.

[0154] ;

[0155] Among them, g i The influence of each feature value obtained from gradient boosting calculation on the prediction of blasting effect, g.i .

[0156] In this embodiment, the number of features involved in predicting blasting effects can be multiple. This embodiment selects only four typical features to illustrate the weight calculation process. The remaining features can be calculated using the same method to determine their corresponding contribution and weight coefficients. These features include uniaxial rock mass strength, mining site stress magnitude, microseismic activity frequency, and drilling rig thrust. The feature contribution values ​​obtained from the gradient boosting calculation are as follows: (Uniaxial rock mass strength) (Magnitude of stress at the mining site) (Frequency of microseismic activity) (Drilling rig propulsion), cumulative The corresponding weight vector is ,in .

[0157] After completing the weight update, the normalized multi-source fusion feature vector Fsync is input into the trained ensemble learning model to obtain the lithological strength index Kv and the breakability index Kp of the target section.

[0158] The calculation results are: Kv = 0.71, Kp = 0.64.

[0159] The aforementioned lithological parameter Kv and breakability index Kp serve as input parameters for the subsequent blasting parameter optimization calculation module, guiding the adaptive adjustment of hole spacing, row spacing, and charge structure.

[0160] The lithology identification results and breakability index inversion results are organized according to borehole, section, or spatial unit, and output to the blasting parameter optimization calculation module in structured data form. The system can generate lithology distribution maps, breakability spatial zoning maps, and corresponding characteristic parameter sets, which serve as the core input conditions for adaptive optimization of blasting parameters and generation of drilling rig collaborative execution commands.

[0161] Through the above-mentioned multi-source data fusion and lithology-breakability inversion process, dynamic perception and quantitative characterization of the geological conditions in front of the mining area are achieved, providing reliable data support and decision-making basis for subsequent optimization calculation of blasting parameters.

[0162] S3, Blasting Parameter Optimization Calculation and Prediction Algorithm: After completing the synchronous fusion of multi-source features and the inversion of lithology and ruptureability, this invention enters the automated calculation, design prediction and full-process optimization of blasting parameters.

[0163] After completing the multi-source data fusion and lithology-breakability inversion, the system enters the adaptive optimization calculation and prediction stage of blasting parameters. This stage uses lithology identification results, breakability index, and construction and equipment constraints as inputs, and achieves zonal optimization and prediction verification of blasting design parameters by constructing blasting parameter-effect response relationships.

[0164] S31, this module uses the fused feature vector F sync and lithological strength index K v Fragility index K p As input, a multi-objective optimization function is used, and the output includes hole spacing S, row spacing B, hole depth H, charge amount Q, and charge structure correction coefficient q. （z） and extended network The optimization function for the multi-objective parameter space of the multiple sets of blasting parameters is as follows:

[0165] ;

[0166] Among them, J frag The model predicts the difference between the block size and the target block size for block size indicators (average block size, oversized block rate); J energy As an indicator of energy utilization rationality, the constraint function avoids unreasonable parameter combinations that result in high energy consumption and low efficiency; where E( E represents the input energy generated per unit of rock breaking corresponding to the current parameter combination. d (k) v k p The required energy is obtained by inversion based on lithological parameters and breakability indices.

[0167] In this embodiment, a set of candidate blasting parameters was selected as follows: S = 1.8 m, B = 1.6 m, H = 8.0 m, Q = 2.82 kg / m³. 3 Dn = 50 ms. Based on the blasting block size prediction model, J... frag The index is 0.12, J energy With an index of 0.026 and weighting coefficients α1=0.6 and α2=0.4, J=0.0824.

[0168] S32, after obtaining multiple sets of candidate blasting parameters and calculating the objective function value, further screening is performed by introducing safety and shaping constraints. Parameter combinations that do not meet the constraints are eliminated, resulting in a feasible solution set that meets the requirements of engineering applications. The introduced safety and shaping constraints are:

[0169] ;

[0170] Among them, J vib The vibration velocity of particles at different distances from the blast center is used to constrain blasting parameters to meet safety control requirements; J shape An index representing the predicted deviation between the profile and the cross-sectional shape after blasting. To allow for relative deviations between over-excavation and under-excavation (e.g., 5%–10%), a parameter matrix is ​​output to constrain over-excavation and under-excavation trends and guide the optimization of molding quality, ultimately achieving the target.

[0171] The allowable over- or under-excavation relative deviation can be dynamically set according to the roadway type, support form, or construction accuracy requirements.

[0172] The above multi-objective optimization function J( It is mainly used for solving and optimizing the blasting parameters. Its value is only used as a comparison and convergence criterion within the algorithm and is not directly used as a basis for engineering implementation decisions.

[0173] In this embodiment, a set of candidate blasting parameters was selected as follows: S = 1.8 m, B = 1.6 m, H = 8.0 m, Q = 2.82 kg / m³. 3 Dn = 50 ms. Design cross-sectional area A design = 9.5 m², predicted post-explosion cross-sectional area A pred = 9.2 m², then: J shape =0.03, take ε=0.10, then J shape = 0.03≤0.10, Predicted blasting vibration velocity V pred = 89.22 cm / s, permissible vibration velocity V limit = 120.49 cm / s, then: J vib = 0.74≤1, which meets the vibration safety constraint requirements.

[0174] Therefore, this parameter combination satisfies the safety and forming constraints and enters the feasible solution set.

[0175] S33, after obtaining multiple sets of candidate blasting parameter schemes that meet the optimization conditions, the system further constructs a comprehensive executability scoring function Score for the blasting scheme based on the prediction model output results. The scoring function is defined as follows:

[0176] Score = λ1·Frag Score +λ2·Energy Score +λ3·Vib Score +λ4·Shape Score；

[0177] Each individual score is obtained by normalizing, classifying, or mapping the prediction results, and is used to transform the blasting effect index into an engineering evaluation value under a unified dimension.

[0178] The system can automatically select the highest-rated solution as the recommended execution plan, and the final execution parameters are issued after confirmation by the engineer.

[0179] In this embodiment, the effectiveness of candidate blasting parameter schemes is evaluated. The blasting parameter combination is: S = 1.8 m, B = 1.6 m, H = 8.0 m, Q = 2.82 kg / m³. 3 For the combination of Dn = 50 ms, based on the calculation results of the blasting effect prediction model, the individual indicators are as follows:

[0180] Block size index: FragScore = 0.88; Energy utilization index: EnergyScore = 0.84; Vibration control index: VibScore = 0.74; Section forming index: ShapeScore = 0.97.

[0181] The scoring weight coefficients are set as follows: λ1=0.4, λ2=0.3, λ3=0.2, λ4=0.1.

[0182] The calculated overall score is 0.849.

[0183] The blasting parameter scheme received a high overall score and is recommended as the implementation scheme.

[0184] In this invention, different weight parameters are applied to different stages such as data fusion, parameter optimization, scheme evaluation and energy modeling. Their domains of application are independent of each other and do not constitute redundant definitions.

[0185] In some other implementations, the specific process is as follows:

[0186] 1) Initial constraints on blasting parameters and construction of the search space

[0187] The system first establishes an initial constraint set of blasting parameters based on the type of mining area, the capabilities of the rock drilling equipment, and the characteristics of the blasting materials. The blasting parameters include at least hole spacing, row spacing, hole depth, charge quantity, charge structure, and delay network configuration.

[0188] The range of values ​​for each parameter is constrained by the following conditions:

[0189] 1. Permissible drilling posture, maximum hole depth, and drilling capacity of the rock drilling equipment;

[0190] 2. Explosive performance parameters and loading process constraints;

[0191] 3. Safety regulations impose limitations on single-hole charge amount, maximum simultaneous charge amount, and vibration control;

[0192] 4. Empirical design intervals corresponding to lithological identification results and breakability index.

[0193] Under the above constraints, the system constructs a feasible search space for blasting parameters, providing boundary conditions for subsequent adaptive optimization.

[0194] 2) Construction of an evaluation index system for blasting effects

[0195] After completing the construction of the blasting parameter search space, the system further establishes a blasting effect evaluation index system to guide parameter optimization and prediction verification. This evaluation index system is used to quantitatively characterize the comprehensive performance of different combinations of blasting parameters in terms of safety, energy efficiency, and forming quality. The indexes are constructed according to the principles of mutual independence, clear physical meaning, and distinction between action stages, avoiding redundant definitions of evaluation dimensions.

[0196] The evaluation indicators for blasting effects include at least the following categories:

[0197] 1. Particle size and crushing effect indicators

[0198] The indicators used to characterize the degree of fragmentation and block size distribution of ore and rock after blasting include, but are not limited to, average block size, large block rate, and block size distribution concentration, and are used to evaluate the impact of blasting on subsequent loading and mining efficiency.

[0199] 2. Energy utilization rationality indicators

[0200] The index is used to evaluate the degree of matching between the blasting input energy and the actual rock breaking effect. It is based on a comprehensive model of unit rock breaking energy consumption, energy redundancy rate and the proportion of underutilized energy, to constrain high-energy-inefficient or unreasonable energy configuration parameter combinations.

[0201] 3. Vibration and safety control indicators

[0202] The indicators used to characterize the disturbance effect of blasting on the surrounding rock and adjacent structures include peak particle velocity, dominant frequency distribution and attenuation characteristics, and are used to constrain blasting parameters to meet safety control requirements.

[0203] 4. Contour shaping and over- and under-excavation control indicators

[0204] The index is used to evaluate the forming quality of the roadway or working face profile after blasting. It constrains the forming quality of blasting parameters by predicting profile deviation, over-excavation and under-excavation, and working face flatness.

[0205] The aforementioned indicators apply to different engineering objectives and constraint dimensions. Their values ​​are not directly equivalent or interchangeable, but rather participate in the parameter search and scheme selection process as independent evaluation terms in the multi-objective optimization function.

[0206] 3) Modeling the relationship between blasting parameters and effect response

[0207] After the evaluation index system is constructed, the system establishes a mapping model between blasting parameters and blasting effects based on historical blasting data, lithology-breakability inversion results, and current construction constraints. This model is used to predict the expected blasting effect of a given combination of blasting parameters under current rock mass conditions.

[0208] The mapping relationship model takes fused feature vectors, lithological indices, and breakability indices as inputs and blasting effect evaluation indices as outputs. Its modeling process includes:

[0209] 1. Construction of parameter-effect samples based on historical construction data;

[0210] 2. Establish regional response models for different lithological sections to avoid the failure of a unified model under strongly heterogeneous conditions;

[0211] 3. Introduce regularization and constraint terms to prevent the model from producing non-physical prediction results in the extreme parameter range.

[0212] Through the above modeling process, the system can quickly predict and compare the expected effects of candidate blasting parameter schemes without conducting actual blasting tests.

[0213] 4) Multi-objective optimization solution and candidate solution generation

[0214] After establishing the parameter-effect response model, the system uses blasting parameters as decision variables and blasting effect evaluation indicators as objective functions to perform a multi-objective optimization process. This optimization process, while satisfying safety and construction constraints, searches for the parameter combination with the optimal overall performance within the feasible parameter space.

[0215] During the optimization process, the system does not pursue an extreme solution for a single objective, but instead generates a set of candidate blasting parameter schemes that satisfy the constraints. Each candidate scheme corresponds to a set of predicted blasting effect indicators, which are used for subsequent comprehensive feasibility evaluation.

[0216] The multi-objective optimization process is only used as a calculation method for generating and screening blasting parameters. Its internal objective function values ​​are only used for scheme comparison and convergence judgment, and are not directly used as the basis for judging engineering implementation.

[0217] 5) Comprehensive feasibility assessment and recommendation of parameter schemes

[0218] After obtaining the set of candidate blasting parameter schemes, the system further conducts a comprehensive feasibility evaluation of each scheme. The evaluation process normalizes evaluation indicators such as block size, energy, safety, and formation, and then weights and summarizes them according to preset weights to form a comprehensive score value under a unified dimension.

[0219] The system ranks the candidate solutions based on the comprehensive scoring results and automatically generates a recommended execution plan. The recommended plan, while meeting all safety and construction constraints, also considers crushing effect, energy utilization efficiency, and forming quality, and is suitable for the current mining area and rock mass conditions.

[0220] The final execution parameters are confirmed and locked by professionals, and serve as the sole input source for generating subsequent drilling execution commands and charge detonation commands.

[0221] 6) Parameter locking and consistency flag

[0222] Once the blasting parameter scheme is confirmed, it is marked as being locked by the system, and a corresponding parameter consistency identifier is generated for full-process verification during drilling, charging, and detonation network configuration.

[0223] During construction, if a deviation is detected between the actual execution parameters and the locked parameters, the system will automatically record the type and magnitude of the deviation and use it as input data for subsequent construction deviation identification and blasting effect feedback.

[0224] S4 outputs drilling and blasting collaborative execution commands. Based on the output results of data fusion analysis and blasting parameter optimization calculation and prediction algorithms, it constructs a collaborative execution command set for drilling rigs and blasting operation units, realizing the automatic mapping and distribution of blasting design parameters to construction equipment control commands.

[0225] After completing the optimization calculation of blasting parameters and locking in the execution plan, the system enters the drilling-blasting collaborative execution and construction process monitoring stage. In this stage, the confirmed blasting parameter plan is the sole control source. Through the parameter-command mapping mechanism, the design parameters are automatically converted into executable commands for drilling, charging, and detonation. During the construction process, the execution status is monitored in real time and deviations are identified to ensure that the design intent is accurately implemented on site.

[0226] S41, Drilling execution command generation module, based on the optimized calculation of the blasting parameter set Θ={S,B,H,Q,q (z) The system automatically generates drilling execution instructions for single-hole and mesh drilling, wherein the instructions include at least:

[0227] 1. Drilling start coordinates and target end coordinates (x, y, z);

[0228] 2. Drilling inclination angle, azimuth angle, and allowable attitude deviation range;

[0229] 3. Single-hole design hole depth H and over-depth compensation amount;

[0230] 4. Control range of drilling speed, thrust, and rotational speed;

[0231] 5. Drilling type identification.

[0232] The drilling execution command is encapsulated using a communication protocol compatible with the drilling rig control system and sent to the drilling rig control terminal through a standard interface to achieve automatic guidance and precise execution of the drilling trajectory.

[0233] S42, Module for Generating Coordinated Commands for Charge and Detonation

[0234] Based on the optimized blasting parameters and lithological inversion results, the system further generates coordinated commands for charge and initiation, including:

[0235] 1. Single-hole charge quantity Q and charge structure segmentation parameters q (z) ;

[0236] 2. Control parameters for charge density, coupling state, and cavity structure;

[0237] 3. Selection of detonation method (electric detonator, electronic detonator, etc.);

[0238] 4. Configuration of delay network Dn and delay time between holes and rows.

[0239] The charging and detonation commands can be interfaced with an electronic detonator programming system or a blasting control terminal to achieve automatic configuration and verification of detonation network parameters.

[0240] S43, Construction process monitoring and deviation identification status feedback interface

[0241] During drilling and charging operations, the following construction status data are collected and transmitted in real time through the drilling rig control system, the measurement while drilling (MWD) system, and the blasting control terminal:

[0242] 1. Actual drilling trajectory, hole depth, hole orientation, and attitude change information;

[0243] 2. The thrust, rotational speed, drilling pressure, torque, and drilling speed during the drilling process;

[0244] 3. In-hole status information (abnormal signals such as stuck drill, deviation, hole collapse, and void indication);

[0245] 4. Actual charge quantity, charge structure execution status, and detonation parameter configuration status.

[0246] The aforementioned monitoring data is encapsulated according to a unified timestamp and spatial coordinate system and transmitted via a status feedback interface.

[0247] S44, Automatic Identification Mechanism for Design-Construction Deviations

[0248] The system compares and analyzes the actual construction data returned from the interface with the original design command parameters, and automatically identifies the following types of deviations:

[0249] 1. Drilling geometric deviations: including hole depth deviation, hole orientation deviation, and hole position offset;

[0250] 2. Deviations in the drilling process: including abnormal drilling energy consumption and sudden changes in drilling footage efficiency;

[0251] 3. Deviations in charge loading: including insufficient charge quantity and inconsistent charge structure;

[0252] 4. Detonation network deviation: including abnormal delay configuration or missing network integrity.

[0253] The deviation identification results are output in the form of quantitative indicators, including deviation amplitude, duration and spatial distribution characteristics.

[0254] S45, Deviation Classification and Handling Strategies

[0255] The system classifies and processes the identified construction deviations according to their type and magnitude:

[0256] For minor deviations within the permissible range, only their characteristic information is recorded and included in the blasting effect feedback data;

[0257] For critical deviations exceeding the set threshold, the system can prompt professionals to conduct manual confirmation and take measures such as supplementary drilling, adjustment of the charge structure, or reconfiguration of detonation parameters as appropriate.

[0258] For serious deviations that pose safety risks, the system can trigger a construction suspension or scheme review mechanism to prevent unreasonable parameters from directly entering the detonation stage.

[0259] Through the above-mentioned deviation identification and handling mechanism, it is ensured that the drilling and blasting construction process is always under control, which can be monitored, evaluated and corrected.

[0260] S5, Blasting Effect Collection and Feedback: Used to quantitatively evaluate the actual effect after blasting construction, and input the evaluation results as parameter feedback signals to the data fusion analysis section and the blasting parameter optimization model, so as to realize the self-learning update and closed-loop optimization of the blasting parameter design model.

[0261] After the blasting operation is completed, the system enters the blasting effect acquisition and feedback update phase. This phase uses post-blast measured data to conduct multi-dimensional quantitative evaluation of the blasting effect, and uses the evaluation results as feedback input for model updates and subsequent adaptive optimization of blasting parameters, thus realizing a closed-loop operation of blasting design-execution-evaluation-update.

[0262] S51, Blasting Effect Data Acquisition: After the blasting is completed, data related to the blasting effect, including but not limited to direct and indirect data, is acquired through monitoring equipment and on-site acquisition methods.

[0263] 1. Blasting vibration monitoring data, including peak particle velocity, dominant frequency characteristics, and attenuation law;

[0264] 2. Post-blast block size distribution data, including large block ratio, average block size, and block size distribution characteristics; obtain the characteristic parameters of ore block size distribution after blasting, including average block size, particle size distribution curve, and large block ratio index, through image recognition, point cloud segmentation, or particle size statistics methods, to evaluate the rationality of blasting energy release.

[0265] 3. Blasting profile data, including over-excavation, under-excavation, and working face flatness indicators;

[0266] 4. Post-explosion energy utilization information, including residual rate and loosening degree assessment results.

[0267] After the above data is registered with a unified timestamp and spatial coordinates, a set of measured features of blasting effects is formed, which is then constructed as a unified blasting deviation feature vector:

[0268]

[0269] ;

[0270] Among them, e frag For block size deviation index; e shape For contour deviation index; e vib For blasting vibration deviation index; e energy As an indicator of blasting energy utilization rate, J is used as the reference in step S2. energy The feedback correction amount of the energy rationality constraint function is used to update the empirical coefficients or correction terms of the unit rock-breaking energy consumption model E(Θ); D pred To predict the average block size for the model; D real The average size of the fragments was measured after the explosion; A pred To predict the post-explosion cross-sectional area for the model; A real v represents the measured cross-sectional area after the explosion. pred Predict peak vibration velocity for the model; v real E represents the measured peak vibration velocity after the blast. in E is the total energy input for the blast. eff For effective rock-breaking energy.

[0271] In this embodiment, the blasting deviation characteristics are calculated based on the predicted blasting effect and the post-blast measured data. Block size deviation index: Model-predicted average block size D pred =0.58m, post-explosion measured average block size D real =0.62m, then: e frag ≈0.065. Cross-sectional forming deviation index; model-predicted cross-sectional area A. pred =9.2 m², measured cross-sectional area A after the explosionreal =9.3m², then: e shape ≈0.011; Blasting vibration deviation index, model predicted peak velocity V pred =89.22cm / s, measured peak vibration velocity V after the explosion real =96.41 mm / s, then: e vib = 0.0746; Energy utilization deviation index, model-predicted unit rock-breaking energy consumption E pred =5.3kJ / m³, measured unit rock-breaking energy consumption E after blast real = 5.0 kJ / m³, then: e energy =0.06.

[0272] Therefore, the blasting deviation feature vector is constructed as follows:

[0273] E blast = {0.065, 0.011, 0.0746, 0.060}.

[0274] Step S5 further includes: quantitative evaluation and index construction of blasting effects. Based on the collected blasting effect data, the system performs multi-index quantitative evaluation of the blasting effects, forming standardized blasting effect evaluation results. The evaluation indicators include at least:

[0275] 1. Deformation index of blasting profile;

[0276] 2. Over-excavation and under-excavation volume and their spatial distribution characteristics;

[0277] 3. Mean size, dispersion, and percentage of large blocks in the ore and rock;

[0278] 4. Blasting vibration intensity and energy utilization characteristics;

[0279] 5. Safety-related abnormal event identification indicators.

[0280] The evaluation indicators are normalized and weighted to form a comprehensive evaluation result that represents the overall effect of the blasting, and are stored in association with the corresponding blasting zone and parameter scheme.

[0281] Step S5 also includes: Design-Execution-Effect Correlation Analysis. The system jointly analyzes the blasting effect evaluation results with blasting design parameters and construction execution process data to identify the correlation between design parameters, construction deviations, and blasting effects. Specifically, this includes:

[0282] 1. Analyze the influence of borehole geometric deviations on blasting profile and block size distribution;

[0283] 2. Analyze the influence of charge structure and detonation network configuration on energy release characteristics;

[0284] 3. Analyze the consistency between the lithology-breakability zoning results and the actual blasting effects.

[0285] Through the above correlation analysis, the key parameter factors that have a dominant influence on the blasting effect and their sensitive ranges are identified.

[0286] S52, Feedback Update and Model Adaptive Correction: Based on the blasting deviation feature vector constructed above, it is used as the feedback input of the blasting parameter design model to characterize the degree of deviation between the design parameter output and the actual blasting effect. It is then introduced into the S2 lithology and breakability inversion algorithm stage to update the prediction weights of each parameter index. This provides support for the next step of inversion.

[0287] The update process does not directly modify the measured or calibrated results of lithology and fragility indices, but is used to correct the model's mapping weights and sensitivity to different features during the inversion process.

[0288] Specifically, based on the blasting effect evaluation results and correlation analysis conclusions, the system updates the lithology identification model, the breakability evaluation model, and the blasting parameter optimization model. The update methods include, but are not limited to:

[0289] 1. Correct the parameters of the lithology-breakability inversion model to improve the accuracy of identifying complex geological conditions;

[0290] 2. Dynamically adjust the recommended rules or optimization weights for blasting parameters to better reflect the actual energy release characteristics;

[0291] 3. The allowable threshold for construction deviations is modified to make the design parameters more feasible on site.

[0292] After the model is updated, the system will automatically apply the updated results to the subsequent blasting parameter design process, so as to achieve continuous adaptive optimization of the blasting scheme.

[0293] Step S5 also includes a mechanism for professional personnel participation and manual verification. During the feedback and update process, the system can display the blasting effect evaluation results, parameter correlation analysis conclusions, and model update suggestions to professional technicians through a visual interface. Professionals can confirm, correct, or reject the update results provided by the system based on their engineering experience. The system records the manual intervention behavior and the final adopted result, and uses it as an additional constraint on the model update process.

[0294] By introducing a professional verification mechanism, the engineering rationality and safety reliability of the blasting parameter update results can be improved while ensuring the system's self-adaptability.

[0295] Through the aforementioned blasting effect feedback and parameter feedback mechanism, the blasting parameter design model can continuously correct its internal mapping relationship during multiple rounds of construction, so that the model output parameters gradually tend to match the actual working conditions on site. When the change of the comprehensive deviation index is less than the preset threshold in multiple rounds of blasting construction, or when the requirements for engineering safety and quality control are met, the current blasting parameter model is considered to have reached a stable state, thereby realizing a closed-loop optimization process of blasting parameter design-construction execution-effect feedback-model update.

[0296] Example 2

[0297] This embodiment 2 provides a system for adaptive design of blasting parameters and collaborative execution of drilling rigs based on multi-source data. This system is used to execute the aforementioned method for adaptive design of blasting parameters and collaborative execution of drilling rigs based on multi-source data, including:

[0298] The data acquisition module is used to acquire multi-source data from the blasting area, including at least drilling measurement data.

[0299] The fusion analysis module is used to perform spatiotemporal synchronous fusion of the multi-source data to generate a fusion feature vector; based on the fusion feature vector, an integrated learning model is used to invert the rock mass mechanical parameters and the breakability index.

[0300] The blasting parameter optimization design module is used to generate a blasting design scheme that includes at least hole mesh parameters and charge parameters based on the rock mass mechanics parameters and through a multi-objective optimization algorithm.

[0301] The drilling and blasting collaborative execution module is used to automatically map the blasting design scheme into executable instructions for the drilling rig and send them to the drilling rig control system to execute the drilling operation;

[0302] The blasting feedback update module is used to acquire post-blast measured effect data, construct a deviation feature vector, and update the model weights used in the fusion analysis module and the blasting parameter design module based on the deviation feature vector to form a closed-loop optimization.

[0303] The fusion analysis module includes:

[0304] The spatiotemporal synchronization unit is used to construct a four-dimensional spatiotemporal index structure and calculate the spatiotemporal weight of each data point based on a Gaussian weighting function.

[0305] The feature mapping unit is used to construct feature value matching for candidate data points, mapping discrete categories into numerical features that can participate in computation, and obtaining a fused feature vector.

[0306] The confidence management unit is used to calculate the variance of the synchronization results through unbiased estimation based on the initial confidence and data point variance of each data source.

[0307] The anomaly suppression unit is used to trigger anomaly marking and compensation processing when the variance of the synchronization result exceeds a threshold, the number of data points participating in the fusion is insufficient, or the weight concentration is too low.

[0308] In summary, this invention provides a method for adaptive design of blasting parameters and collaborative execution of drilling rigs based on multi-source data, relating to the field of intelligent mining technology. By introducing multi-source raw data such as measurement while drilling and geophysical exploration as input, an adaptive fusion algorithm incorporating spatiotemporal synchronous weighting, confidence management, and anomaly suppression is used for data processing. Then, based on ensemble learning, lithology and breakability indices are inverted, and optimal blasting parameters are automatically generated through a mechanism that integrates multi-objective optimization and comprehensive executability scoring. During the construction phase, the system automatically identifies design-execution deviations through real-time monitoring and transmits the status back. Finally, the post-blasting measured results are constructed as a deviation feature vector and fed back to the design model, enabling self-learning updates of the model weights. This allows the entire blasting process to form a quantifiable, verifiable, and iterative closed loop, completely transforming traditional blasting design, which relies on manual experience, into an intelligent scientific decision-making process driven by data and algorithms.

[0309] It should be noted that the present invention is not limited to the above-described embodiments. The above embodiments are merely examples, and any embodiments that have the same structure and perform the same effects as the technical concept within the scope of the present invention are included within the scope of the present invention. Furthermore, various modifications that can be conceived by those skilled in the art to the embodiments, and other ways of constructing by combining some of the constituent elements of the embodiments, without departing from the spirit of the present invention, are also included within the scope of the present invention.

Claims

1. A method for adaptive design of blasting parameters and collaborative execution of drilling rigs based on multi-source data, characterized in that, Includes the following steps: S1, Data Acquisition and Input: Acquire multi-source data of the blasting area, including drilling measurement data, geophysical data, 3D scanning data, mining site geological data, and historical construction record data; S2, Data Fusion Analysis: The multi-source data are spatiotemporally fused to generate a fused feature vector; based on the fused feature vector, an integrated learning model is used to invert the rock mass mechanical parameters and the breakability index. S3, Blasting parameter optimization design: Based on the rock mass mechanics parameters and breakability index, a blasting design scheme including at least hole mesh parameters and charge parameters is generated through a multi-objective optimization algorithm; S4, Drilling and blasting coordinated execution: The blasting parameter set is automatically mapped into drilling rig control commands and charge initiation and detonation coordinated commands and issued for execution; Real-time collection of construction measurement data, identification of design-construction deviations and dynamic correction of the charge initiation and detonation coordinated commands; S5, Blasting effect feedback update: Obtain post-blast measured effect data, quantify the deviation between blasting design parameters and actual blasting effect, and use the deviation feature vector as feedback input to introduce data fusion analysis and blasting parameter optimization design process. Update the parameters or feature weights of the lithology inversion model and blasting effect prediction model in step S2, thereby providing a basis for the blasting parameter design in the subsequent step S3, forming a closed-loop optimization.

2. The method for adaptive design of blasting parameters and collaborative execution of drilling rigs based on multi-source data as described in claim 1, characterized in that, In step S2, the data fusion analysis employs a multi-source fusion algorithm based on adaptive weight adjustment, specifically including: S21, Spatiotemporal Synchronization Model Construction: Construct a unified four-dimensional spatiotemporal index ID=f(x,y,z,t), and establish a nearest neighbor matching structure for multi-source data using a kd-tree structure combined with timestamp sorting. This structure is applied to each data point D in the multi-source data. j Spatial distance and time difference are calculated, and spatiotemporal weights are calculated based on a Gaussian weighting function to achieve spatiotemporal synchronization of multi-source data; wherein, the calculation formula for the spatiotemporal weights is as follows: ； Among them, TS j For spatiotemporal weights; D j For the data points participating in the matching; d s d represents spatial distance; t For time difference; and These are the spatial and temporal scale parameters, respectively; C j The initial confidence level for candidate data sources; S22, Feature Mapping: Constructing Feature Value Matching F for Candidate Data Points j An embedding transformation method is used to map discrete categories into computationally computable numerical features. After embedding, all categorical variables form a uniform vector space, which, together with continuous variables, constitutes a multi-source feature vector, ultimately yielding a fused feature vector F. sync The formula is as follows: ； S23, Confidence Management: Based on the weight distribution of each data source participating in the fusion and the variance of candidate data points, the variance σ of the synchronization result is calculated through unbiased estimation. sync A confidence management mechanism is constructed; wherein, the variance σ sync The calculation formula is as follows: ； S24, Anomaly Suppression: When the variance of the synchronization result exceeds the threshold, the number of data points participating in the fusion is insufficient, or the weight concentration is too low, the anomaly suppression mechanism is triggered to mark the synchronization anomaly unit and perform compensation processing.

3. The method for adaptive design of blasting parameters and collaborative execution of drilling rigs based on multi-source data according to claim 2, characterized in that, In step S2, the specific process of lithological classification and breakability inversion in the data fusion analysis is as follows: S25 employs a combination of an ensemble learning model and an optimized classification tree method, inputting the cleaned multi-source feature vectors into the ensemble learning model for inverting the lithological strength index K. v Fragility index K p The ensemble learning model, while completing the lithological parameter inversion, further constructs a blasting effect prediction model to establish a blasting parameter set. The mapping relationship between block size distribution, vibration response and cross-section forming results; S251, the ensemble learning backbone model adopts XGBoost, GBDT or LightGBM gradient boosting ensemble model, and constructs multiple weak learning trees through boosting iteration to achieve high-precision classification of lithology and continuous prediction of breakability index. S252, optimize the classification tree auxiliary model as the structural constraint and assistant of the backbone model. Select CART classification tree or cost-sensitive classification tree to perform structured hierarchical discrimination of the response of key lithology categories in blasting. S253, a weight control and feature fusion mechanism, introduces a confidence-based weight control and multi-layer feature fusion mechanism during training and inference. The model assigns weight factors that are updated in real time to each feature based on its contribution to the explosion effect prediction. This achieves adaptive adjustment to strengthen high-contribution features and weaken low-contribution features; among which, the weighting factor The calculation formula is as follows: ； Among them, g i The influence of each feature value obtained from gradient boosting calculation on the prediction of blasting effect.

4. The method for adaptive design of blasting parameters and collaborative execution of drilling rigs based on multi-source data according to claim 3, characterized in that, In step S3, the optimization design of the blasting parameters specifically includes: S31, using the fused feature vector F sync and lithological strength index K v Fragility index K p As input, a multi-objective optimization function is used, and the output includes hole spacing S, row spacing B, hole depth H, charge amount Q, and charge structure correction coefficient q. （z） and extended network The optimization function for the multi-objective parameter space of the multiple sets of blasting parameters is as follows: ； Among them, J frag The block size index model predicts the difference between the block size and the target block size in the project; J energy As an indicator of energy utilization rationality, the constraint function avoids unreasonable parameter combinations that result in high energy consumption and low efficiency; E( E represents the input energy generated per unit of rock breaking corresponding to the current parameter combination. d (k) v k p The required energy is obtained by inversion based on lithological parameters and breakability indices. , Let be the weight coefficients of the objective function, and satisfy . Used to balance block size control and energy utilization efficiency; , is the weighting coefficient for the i-th spatial unit or sampling region, which is determined based on the volume ratio of the sampling region or the grade of the ore. This is a set of blasting parameters.

5. The method for adaptive design of blasting parameters and collaborative execution of drilling rigs based on multi-source data according to claim 4, characterized in that, In step S3, the optimization design of the blasting parameters further includes: S32 introduces safety and molding constraints: ； Among them, J vib The vibration velocity of particles at different distances from the blast center is used to constrain blasting parameters to meet safety control requirements; J shape An index representing the predicted deviation between the profile and the cross-sectional shape after blasting; To allow for relative deviations between over- and under-excavation, used to constrain over- and under-excavation trends and guide the optimization of forming quality; A pred The predicted cross-sectional profile and area after blasting; A design This represents the design values ​​for the cross-sectional profile and area after blasting; v pred This is the predicted value of particle vibration; v limit These are the safety limits for particle vibration. S33, after obtaining multiple sets of candidate blasting parameter schemes that meet the optimization conditions, based on the output results of the blasting effect prediction model described in step S25, a comprehensive executability scoring function Score for the blasting scheme is further constructed. The scoring function Score is defined as: Score =λ1·Frag Score +λ2·Energy Score +λ3·Vib Score +λ4·Shape Score ; Among them, Frag Score This is an evaluation value constructed based on the deviation between the predicted block size distribution and the target block size according to the current blasting scheme. The evaluation value is calculated based on the average block diameter, the rate of oversized blocks, or the difference in the block size distribution curve, and is used to reflect the effectiveness of block size control. (Energy) Score This is an evaluation value constructed based on the degree of matching between the unit rock-breaking energy consumption of the current blasting scheme and the theoretical optimal energy consumption. The evaluation value is based on the actual input energy and the theoretical energy consumption E. d The deviation is calculated and used to measure energy utilization efficiency; Vib Score This is an evaluation value constructed based on the relationship between predicted vibration velocity and safety threshold according to the current blasting scheme. The evaluation value is based on the predicted particle vibration velocity v. pred With the allowable vibration velocity v limit The deviation is calculated and used to assess blasting safety; Shape Score The evaluation value is constructed based on the deviation between the post-blast cross-sectional profile and area of ​​the current blasting scheme and the designed cross-section. This evaluation value is calculated according to the cross-sectional area deviation rate or profile fitting error and is used to reflect the forming quality; λ 1、 λ 2、 λ 3、 λ4 represents the weighting coefficients for each evaluation value, satisfying: It is used to achieve a comprehensive optimization goal that balances block size control, energy utilization, safety, and molding quality; Each individual score is obtained by normalizing, classifying, or mapping the prediction results, and is used to transform the blasting effect index into an engineering evaluation value under a unified dimension.

6. The method for adaptive design of blasting parameters and collaborative execution of drilling rigs based on multi-source data according to claim 4, characterized in that, In step S4, the drilling and blasting coordinated execution specifically includes: S41, Drilling execution command generation: Based on the blasting parameter set Θ={S,B,H,Q,q} (z) ,Dn}, automatically generate drilling execution instructions for single hole level and hole network level; the instructions include at least the drilling start and end coordinates, drilling inclination angle and azimuth angle and allowable attitude deviation range, single hole design depth and over-depth compensation amount, drilling speed and propulsion force and rotation speed control range, and drilling type identifier. S42, Generation of Charge and Initiation Coordination Commands: Based on the optimization results of blasting parameters and lithology inversion results, further generate charge and initiation coordination commands, including single-hole charge amount Q and charge structure segmentation parameters q. (z) , charge density and coupling state and cavity structure control parameters, initiation method selection, delay network Dn and inter-hole row delay time configuration; S43, collects and transmits construction status data: Through the drilling rig control system, drilling measurement system and blasting control terminal, it collects and transmits construction status data in real time, including actual borehole trajectory and depth, borehole direction and attitude change information, propulsion force and rotation speed, drilling pressure and torque and footage speed during drilling, abnormal signals in the borehole, actual charge quantity and charge structure execution status and detonation parameter configuration status; compares and analyzes the actual construction data with the original design command parameters, automatically identifies borehole geometric deviation, drilling process deviation, charge execution deviation, and detonation network deviation, and outputs quantitative deviation amplitude, duration and spatial distribution characteristics; S44, Automatic Design-Construction Deviation Identification: Compares and analyzes construction status data with original design command parameters to automatically identify deviations, including borehole geometric deviations, drilling process deviations, charge execution deviations, and detonation network deviations.

7. The method for adaptive design of blasting parameters and collaborative execution of drilling rigs based on multi-source data according to claim 6, characterized in that, In step S5, the blasting effect feedback update is used to quantitatively evaluate the actual effect after the blasting operation is completed, and the evaluation result is used as a parameter feedback signal to be input into the data fusion analysis and blasting parameter optimization, so as to realize the self-learning update and closed-loop optimization of the blasting parameter design model, specifically including: S51. After the blasting is completed, collect the measured data of the blasting effect, including blasting vibration monitoring data, post-blast block size distribution data, blasting profile data, and post-blast energy utilization information. S52, after registering the collected data with a unified timestamp and spatial coordinates, forms a set of measured features of the blasting effect, and constructs a unified blasting deviation feature vector E. blast , means as follows: ； ； Among them, e frag For block size deviation index; e shape For contour deviation index; e vib For blasting vibration deviation index; e energy As an indicator of blasting energy utilization rate, J is used as the reference in step S2. energy The feedback correction amount of the energy rationality constraint function is used to update the empirical coefficients or correction terms of the unit rock-breaking energy consumption model E(Θ); D pred To predict the average block size for the model; D real The average size of the fragments was measured after the explosion; A pred To predict the post-explosion cross-sectional area for the model; A real v represents the measured cross-sectional area after the explosion. pred To predict peak vibration velocity for the model; v real E represents the measured peak vibration velocity after the blast. in E is the total energy input for the blast. eff For effective rock-breaking energy; S53, the blasting deviation feature vector is introduced as a feedback input into the lithology inversion stage to characterize the degree of deviation between the design parameter output and the actual blasting effect. This vector is then introduced into the S2 lithology and breakability inversion algorithm stage to update the prediction weights of each parameter index. This provides support for the next step of inversion.

8. The method for adaptive design of blasting parameters and collaborative execution of drilling rigs based on multi-source data according to claim 2, characterized in that, In the feature mapping described in step S22, one-hot encoding or embedding transformation is used to map discrete categories into numerical features that can participate in the calculation. One-hot encoding is suitable for variables with a small number of categories and clear mutual exclusion relationships. Embedding transformation embedding vectors are suitable for situations with a large number of categories, similarity between categories, or the need to express continuous implicit semantics. Step S24, the specific process of anomaly suppression, is as follows: If any of the following conditions occur, anomaly identification is triggered and the unit is marked as a "synchronization anomaly unit": Case 1, synchronous variance σ sync Greater than the set threshold σ thr ; Scenario 2, Number N of data points participating in the fusion <N min ; Case 3: The weight concentration is too low, Max( ) / Sum( )< ; Case 4: Spatial distance or time difference exceeds the set tolerance range, d s >R tol or d t >T tol This could lead to potential cross-hole or cross-segment errors; Furthermore, synchronization control is performed on the synchronization anomaly units, including: reducing confidence weight, triggering compensation retrieval, extending the time buffer, and calling interpolation predictions of similar data, to ensure the availability of the final synchronization results and avoid interference from mismatches of different spatiotemporal data on subsequent lithology inversion and blasting parameter design.

9. A system for adaptive design of blasting parameters and collaborative execution of drilling rigs based on multi-source data, characterized in that, The method for implementing the adaptive design of blasting parameters based on multi-source data and the collaborative execution of drilling rigs according to any one of claims 1 to 8 includes: The data acquisition module is used to acquire multi-source data from the blasting area, including at least drilling measurement data. The fusion analysis module is used to perform spatiotemporal synchronous fusion of the multi-source data to generate a fusion feature vector; based on the fusion feature vector, an integrated learning model is used to invert the rock mass mechanical parameters and the breakability index. The blasting parameter optimization design module is used to generate a blasting design scheme that includes at least hole mesh parameters and charge parameters based on the rock mass mechanics parameters and through a multi-objective optimization algorithm. The drilling and blasting collaborative execution module is used to automatically map the blasting design scheme into executable instructions for the drilling rig and send them to the drilling rig control system to execute the drilling operation; The blasting feedback update module is used to acquire post-blast measured effect data, construct a deviation feature vector, and update the model weights used in the fusion analysis module and the blasting parameter design module based on the deviation feature vector to form a closed-loop optimization.

10. The adaptive design system for blasting parameters based on multi-source data and the collaborative execution system for drilling rigs according to claim 9, characterized in that, The fusion analysis module includes: The spatiotemporal synchronization unit is used to construct a four-dimensional spatiotemporal index structure and calculates the spatiotemporal weight of each data point based on a Gaussian weighting function. The feature mapping unit is used to construct feature value matching for candidate data points, mapping discrete categories into numerical features that can participate in computation, and obtaining a fused feature vector. The confidence management unit is used to calculate the variance of the synchronization results through unbiased estimation based on the initial confidence and data point variance of each data source. The anomaly suppression unit is used to trigger anomaly marking and compensation processing when the variance of the synchronization result exceeds a threshold, the number of data points participating in the fusion is insufficient, or the weight concentration is too low.

Images