A method and system for analyzing open pit mine blast data

Abstract

The present application aims at the problems existing in the current open pit mining blast data evaluation system technology, and provides an open pit mining blast data analysis method and system, belonging to the field of blasting technology. It comprises: collecting data of the open pit mining blast area; preprocessing the collected data; extracting characteristic parameters representing the blasting effect from the preprocessed data and the data after blasting; constructing a blasting effect evaluation model based on the characteristic parameters, calculating the comprehensive blasting effect score; and generating a blasting parameter adjustment suggestion if the score is lower than the threshold. The present application can comprehensively quantify the blasting quality, environmental response and process stability, and provides a scientific quantitative basis for objectively evaluating the blasting effect. Moreover, it can automatically generate a quantifiable and reproducible adjustment scheme according to the optimization rules mined from historical data, effectively reducing the explosive consumption, reducing the boulder rate fluctuation, stabilizing the vibration hazard, and significantly improving the intelligent level of blasting operation.

Description

Technical Field

[0001] This invention belongs to the field of blasting technology, and specifically relates to a method and system for analyzing blasting data in open-pit mines. Background Technology

[0002] In open-pit mining, blasting is a crucial rock-breaking process, and its effectiveness directly impacts the efficiency of subsequent loading, transportation, and crushing. Compared to underground mines and foundation pits, open-pit blasting faces more complex free-face conditions, greater environmental disturbance risks, and the pressure of continuous production. Its special requirements are mainly reflected in strict control of blasting vibration and flyrock.

[0003] In traditional blasting operations, the acquisition of borehole parameters, rock mechanics parameters, and blasting vibration data is often independent, lacking unified data acquisition standards and a systematic data management platform. For example, parameters such as borehole charge amount and depth are usually recorded manually by on-site technicians, which is prone to errors and omissions and makes it difficult to establish a quantitative correlation with subsequent blasting effects. Rock mechanics parameters are mostly obtained from laboratory tests or geological survey reports, with long update cycles, failing to reflect the dynamic changes in local lithology in the blasting area. Although portable vibration meters are widely used for blasting vibration monitoring, the monitoring data is mostly stored independently on a per-blast basis, lacking effective integration and analysis with parameters such as borehole layout and charge structure. In addition, existing blasting effect evaluation methods usually only focus on a single indicator in block size distribution or vibration peak value, such as obtaining average block size through sieving or photography, or determining whether it exceeds the standard based on vibration monitoring results, without establishing a multi-dimensional comprehensive evaluation system covering block size, vibration, and charge process stability.

[0004] The aforementioned technical defects make it difficult for blasting designs to adapt to the complex and ever-changing rock mass conditions of open-pit mines, resulting in problems such as high explosive consumption per unit, large fluctuations in the proportion of large blocks, and unstable control of vibration hazards.

[0005] Therefore, it is necessary to provide a multi-dimensional comprehensive evaluation system suitable for blasting in open-pit mines in order to improve the overall blasting quality and effect in open-pit mines. Summary of the Invention

[0006] The purpose of this invention is to address the problems existing in current open-pit mine blasting data evaluation systems by providing a method and system for analyzing open-pit mine blasting data. This invention overcomes the limitations of existing technologies that focus only on a single indicator, comprehensively quantifying blasting quality, environmental response, and process stability, providing a scientific quantitative basis for objectively evaluating blasting effects. Furthermore, based on optimization rules mined from historical data, it can automatically generate quantifiable and reproducible adjustment schemes, effectively reducing explosive consumption per unit, minimizing large-block fluctuations, stabilizing vibration hazards, and significantly improving the intelligence level of blasting operations.

[0007] One of the technical solutions of the present invention is a method for analyzing blasting data in open-pit mines, comprising the following steps:

[0008] S1. Collect borehole parameter data, rock mass physical and mechanical parameter data, and blasting vibration monitoring data in the blasting area of ​​the open-pit mine.

[0009] S2. Preprocess the collected data, remove outliers and fill in missing values;

[0010] S3. Extract characteristic parameters representing the blasting effect from the pre-processed data and the post-blasting data. The characteristic parameters include at least the characteristic value of blasting block size distribution, the blasting vibration attenuation index, and the uniformity index of charge density in the blast hole.

[0011] S4. Construct a blasting effect evaluation model based on the aforementioned feature parameters, and calculate the comprehensive blasting effect score;

[0012] S5. Compare the comprehensive blasting effect score with a preset threshold. If the score is lower than the threshold, generate a blasting parameter adjustment suggestion.

[0013] Furthermore, the above-mentioned open-pit mine blasting data analysis method also includes S6: outputting the blasting parameter adjustment suggestions to the data receiving terminal and / or blasting operation control system.

[0014] Furthermore, in the above-mentioned open-pit mine blasting data analysis method, in S1, the borehole parameter data includes the coordinate position, diameter, depth, dip angle, charge structure, detonation sequence, and charge amount per borehole for each borehole; the rock mass physical and mechanical parameter data includes the rock's compressive strength, tensile strength, elastic modulus, Poisson's ratio, and the grading value of the degree of joint and fracture development; the blasting vibration monitoring data includes vibration waveforms collected at different distances from the blasting point, peak particle velocity, dominant frequency, and vibration duration.

[0015] Furthermore, in the above-mentioned open-pit mine blasting data analysis method, the preprocessing step in S2 includes: using an algorithm based on local outlier factors to identify and remove outliers in the borehole parameter data;

[0016] Missing data is filled by a weighted average of parameters of the same type from adjacent boreholes, where the weighting coefficient is determined inversely to the borehole spacing.

[0017] Furthermore, in the above-mentioned open-pit mine blasting data analysis method, in step S3, the characteristic values ​​of blasted block size distribution include characteristic particle size values ​​and a uniformity coefficient characterizing the uniformity of block size distribution. The extraction method is as follows:

[0018] Images of the blasted piles were collected, and the rock block outlines were identified using an edge detection algorithm. The volume percentage of rock blocks of each grain size was calculated, and a block size distribution curve was obtained by fitting the curve. The characteristic grain size value and uniformity coefficient were then extracted from the curve.

[0019] Furthermore, in the above-mentioned open-pit mine blasting data analysis method, the method for extracting the blasting vibration attenuation index in step S3 is as follows:

[0020] Based on the peak particle velocity data collected by vibration monitoring sensors placed at different distances from the blasting point, the vibration attenuation law is fitted by linear regression method to obtain the blasting vibration attenuation index.

[0021] Furthermore, in the above-mentioned open-pit mine blasting data analysis method, the calculation method for the uniformity index of the charge density in the blast holes in step S3 is as follows:

[0022] Calculate the deviation rate between the actual charge density and the designed charge density for each borehole, and statistically analyze the root mean square value of the deviation rates for all boreholes. Use the reciprocal of this root mean square value as the uniformity index; the larger the uniformity index, the better the charge uniformity.

[0023] Furthermore, in the above-mentioned open-pit mine blasting data analysis method, in S4, the blasting effect evaluation model adopts a weighted summation method, wherein the weight coefficients of each characteristic parameter are dynamically determined based on historical blasting data through the entropy weight method or the analytic hierarchy process, and the comprehensive blasting effect score is the sum of the standardized values ​​of each characteristic parameter obtained after processing and their corresponding weights.

[0024] Furthermore, in the above-mentioned open-pit mine blasting data analysis method, step S5, the step of generating blasting parameter adjustment suggestions, includes:

[0025] The current blasting parameters and the comprehensive blasting effect score are input into a pre-trained decision tree model. The decision tree model outputs the type of blasting parameters that need to be adjusted, as well as the adjustment direction and adjustment range, according to the node splitting rules.

[0026] Furthermore, in the above-mentioned open-pit mine blasting data analysis method, in step S5, if the comprehensive blasting effect score of the generated blasting parameter adjustment suggestion is still lower than the preset threshold, then the matching parameters are retrieved from the blasting parameter knowledge base as the blasting parameter adjustment suggestion output; wherein the blasting parameter knowledge base is preset and stores the optimal blasting parameter range under different rock mass conditions.

[0027] The second technical solution of the present invention is an open-pit mine blasting data analysis system, used to implement the above-mentioned open-pit mine blasting data analysis method, comprising:

[0028] The data acquisition module is used to acquire borehole parameter data, rock mass physical and mechanical parameter data, and blasting vibration monitoring data;

[0029] The data preprocessing module is used to clean, denoise, and imput the collected data;

[0030] The feature extraction module is used to extract feature parameters from the data, including at least the characteristic value of blasting block size distribution, the blasting vibration attenuation index, and the uniformity index of charge density in the blast hole;

[0031] The evaluation model module is used to calculate the comprehensive blasting effect score based on the feature parameters;

[0032] The optimization decision-making module is used to generate suggestions for adjusting blasting parameters based on the comprehensive blasting effect score;

[0033] The output module is used to send adjustment suggestions in a visual form or in the form of control commands to the data receiving terminal and / or the execution unit of the blasting operation.

[0034] Furthermore, the data acquisition module of the aforementioned open-pit mine blasting data analysis system also includes an unmanned aerial vehicle (UAV) image acquisition unit for acquiring images of blast piles, a distributed vibration sensor network unit for acquiring peak particle velocities during blasting, and a blast hole measurement unit for acquiring the mass of explosives and the length of the charge section in the blast hole. Each unit aggregates the data to the data aggregation unit of the feature extraction module.

[0035] Furthermore, in the aforementioned open-pit mine blasting data analysis system, the optimization decision module has a built-in blasting parameter knowledge base. This knowledge base stores the optimal blasting parameter ranges under different rock mass conditions. When the comprehensive blasting effect score of the blasting parameter adjustment suggestion is compared with a preset threshold, if the score is lower than the threshold, the matching parameters are retrieved from the knowledge base as the blasting parameter adjustment suggestion output.

[0036] Compared with the prior art, the beneficial effects of the present invention are:

[0037] The method of this invention collects borehole parameters, rock mechanics parameters, and blasting vibration data, and uses a local outlier factor algorithm to automatically identify and remove abnormal charge data and borehole parameter data. It fills in missing values ​​using an inverse distance weighting method, optimizes the traditional manual recording method, strengthens the correlation between isolated data, provides a high-quality and complete data foundation for subsequent feature extraction, and significantly improves the reliability and automation of data analysis.

[0038] Meanwhile, characteristic particle size and uniformity coefficient are extracted from the blast pile size distribution, attenuation index is extracted from vibration monitoring, and charge density uniformity index is calculated from charge records. The entropy weight method is used to dynamically determine the weight of each index, and the weighted calculation yields a comprehensive blasting effect score. This model breaks through the limitation of existing technologies that only focus on a single index, and can comprehensively quantify blasting quality, environmental response and process stability, providing a scientific quantitative basis for objectively evaluating blasting effects.

[0039] Furthermore, based on the decision tree model, using the current blasting parameters and comprehensive score as input, specific parameter adjustment suggestions are output and sent directly to the execution unit of the blasting operation (i.e., the detonator and the charging vehicle) through the output module, forming a closed-loop control. Compared with the traditional method of relying on experience for rough parameter adjustment, this invention can automatically generate quantifiable and reproducible adjustment schemes based on optimization rules mined from historical data, effectively reducing explosive consumption per unit, reducing large block rate fluctuations, stabilizing vibration hazards, and significantly improving the intelligence level of blasting operations. Attached Figure Description

[0040] Figure 1 This is a schematic flowchart of the method of the present invention;

[0041] Figure 2 This is a graph showing the distribution of the explosive block size in this invention.

[0042] Figure 3 This is a diagram showing the attenuation curve of the blasting vibration of the present invention;

[0043] Figure 4 This is a comparison chart of the overall blasting effect score before and after the optimization of blasting parameters in this invention. Detailed Implementation

[0044] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0045] Example

[0046] like Figures 1-4 As shown, the embodiment provides a technical solution: a method for analyzing blasting data in open-pit mines, comprising:

[0047] In the blasting area of ​​the open-pit mine, multiple data acquisition points are set up. The borehole parameter data includes the coordinate position, diameter, depth, dip angle, charge structure, detonation sequence, and charge amount per borehole for each borehole. The rock mass physical and mechanical parameter data includes the compressive strength, tensile strength, elastic modulus, Poisson's ratio, and the graded values ​​of joint and fracture development. The blasting vibration monitoring data is collected by three-component seismic detectors placed at different distances from the blast source, recording the vibration waveform generated by the blast, and extracting the peak particle velocity, dominant frequency, and vibration duration. In this scheme, the acquisition frequency is set according to the blasting frequency, and data is collected before, during, and after each blasting operation.

[0048] After data acquisition, a preprocessing step is initiated. Due to the complex field environment, the acquired data often contains outliers and missing values. This method employs an algorithm based on local outlier factors to identify outliers. Specifically, for the charge amount data of each borehole, the ratio of its local reachable density to that of neighboring boreholes is calculated. If this ratio exceeds a preset upper or lower limit, it is identified as an outlier and removed. After removing outliers, missing data points are filled using a weighted average of similar parameters from adjacent boreholes. The weighting coefficient is calculated as follows: Let the borehole number where the missing data point is located be i, the adjacent borehole number be j, and the spatial distance between borehole i and borehole j be dij, then the weighting coefficient aij = The summation iterates through all adjacent boreholes, so the closer the boreholes are, the greater their data contribution.

[0049] For example: Suppose that in a certain open-pit mine blasting operation, a total of 10 blast holes are arranged. According to the blasting design, the designed charge of each blast hole is 180 kg. However, due to changes in rock conditions and errors in the charging equipment, the actual charge of each blast hole is recorded as follows (in order of blast hole number 1-10): 178 kg, 182 kg, 176 kg, 181 kg, 179 kg, 45 kg, 183 kg, 177 kg, 180 kg, and 184 kg. It can be found that the charge of blast hole number 6, 45 kg, deviates significantly from the other values. This may be due to a malfunction in the charging equipment or a recording error. It is necessary to determine whether it is an outlier and decide whether to remove it. This method uses the local outlier factor algorithm to calculate the local reachable density ratio of the charge of each blast hole to that of its neighboring blast holes. Let K=3 be the number of adjacent boreholes. Since the boreholes are linearly arranged in space, the difference in charge is mainly reflected in numerical values. Therefore, the absolute difference in charge values ​​can be directly used as the distance measure. First, calculate the local reachability density of borehole No. 6 (charge 45 kg). Find the three boreholes closest to 45 kg: borehole No. 9 (difference 135), borehole No. 1 (difference 133), and borehole No. 8 (difference 132). The charge values ​​of these three adjacent boreholes are 177, 178, and 180 kg, respectively. Before calculating the local reachability density of each adjacent borehole, we need to find the three nearest neighbors of each adjacent borehole. Taking borehole No. 9 as an example, its three nearest neighbors are borehole No. 10. The difference between the 9th borehole and its three nearest neighbors is 4, 2, 2, and 2. These neighbors have very small differences from the value of 180 kg. The average reachable distance from the 9th borehole to its three nearest neighbors is approximately (2+2+4) / 3 = 2.67. Its local reachability density is the reciprocal of this average distance, approximately 0.375. Similarly, the three nearest neighbors of the 1st borehole are the 8th borehole, with a difference of 1, 9th borehole, a difference of 2, and 2nd borehole, and a difference of 4. The average distance is approximately (1+2+4) / 3 = 2.33, and the local reachability density is approximately 0.429. The three nearest neighbors of the 8th borehole are the 1st borehole, with a difference of 1, 9th borehole, a difference of 1, a difference of 3, and 3rd borehole, and a difference of 1. The average distance is approximately (1+3+1) / 3 = 1.67, and the local reachability density is approximately 0.599. The average local reachability density of these three adjacent boreholes is approximately (0.375+0.429+0.599) / 3 = 0.468. Now calculate the local achievable density of borehole No. 6 itself.The average reachable distance from borehole 6 to its three neighbors (177, 178, 180) is (132+133+135) / 3≈133.33, and its local reachable density is approximately 1 / 133.33=0.0075. Finally, the local reachable density ratio is calculated, which is the average density of the neighbors divided by its own density: 0.468 / 0.0075=62.4. This ratio far exceeds the preset upper limit of 3.0. Therefore, the charge amount of borehole 6 is judged as an outlier and is removed. In this way, the algorithm can automatically identify outliers where the charge amount of a single borehole deviates significantly from the group, ensuring that subsequent feature extraction and evaluation models are not affected by erroneous data.

[0050] For example, if the depth record of a blast hole is missing, suppose there are 5 blast holes arranged in a straight line at equal intervals on an open-pit mine bench, with an adjacent blast hole spacing of 3 meters. The blast holes are numbered sequentially from 1 to 5, and their actual measured depths are: 12.0 meters, 12.2 meters, missing, 12.1 meters, and 11.9 meters, respectively. We need to fill in the depth of blast hole number 3. We select blast holes number 2 and 4, which are adjacent to blast hole number 3, as references. The distance between blast hole number 3 and blast hole number 2 is 3 meters, and the distance between blast hole number 3 and blast hole number 4 is also 3 meters. Calculate the reciprocals of the two distances: 1 / 3 ≈ 0.3333, 1 / 3 ≈ 0.3333, summing them to 0.6667. Then calculate the weighting coefficients: a32 = 0.3333 / 0.6667 = 0.5, aij = 0.3333 / 0.6667 = 0.5. The depth of borehole No. 2 is 12.2 meters, and the depth of borehole No. 4 is 12.1 meters. Therefore, the depth fill value of borehole No. 3 is: 0.5 × 12.2 + 0.5 × 12.1 = 12.15 meters. If the borehole spacing is unequal, for example, the distance between borehole No. 3 and No. 2 is 4 meters, and the distance between No. 3 and No. 4 is 2 meters, then the reciprocals of the distances are 0.25 and 0.5 respectively, summing to 0.75. The weighting factors are 0.25 / 0.75 ≈ 0.333 and 0. 0.5 / 0.75≈0.667, the filled value is 0.333×12.2+0.667×12.1≈4.06+8.07=12.13 meters. It can be seen that the No. 4 blast hole, which is closer, has received a greater weight. This filling method utilizes the spatial continuity and correlation of rock mass properties in the blasting area. Compared with directly taking the average or median, it can more realistically restore the local features of the missing data location, thereby ensuring the accuracy of subsequent feature extraction.

[0051] Feature parameters characterizing the blasting effect are extracted from both preprocessed and post-blasting data. Specifically, firstly, the block size distribution feature values ​​are extracted. When acquiring images of the blasted pile, a drone equipped with a high-resolution camera is used. After blasting, the drone flies over the blast pile area and captures an image sequence with an overlap rate of no less than 70% along a preset flight path. A 3D point cloud model of the blast pile is generated using motion reconstruction structure and multi-view stereo vision algorithms. Individual rock blocks are then segmented from the point cloud model. An edge detection algorithm is used to extract the contour of each rock block on a 2D projection plane. The equivalent diameter is calculated based on the contour, and the volume of the rock block is estimated. All rock blocks are divided into several grain sizes according to their equivalent diameter, such as less than 50 mm, 50 to 100 mm, 100 to 200 mm, 200 to 400 mm, and greater than 400 mm. The total volume of rock blocks within each grain size range is calculated, and the volume percentage of rock blocks within each grain size range is statistically analyzed to obtain the cumulative volume distribution curve, such as... Figure 2 As shown, the curve is fitted using the Rosin-Rammler distribution function, and the characteristic particle size value and the uniformity coefficient characterizing the uniformity of the block size distribution are extracted from the curve. The characteristic particle size value is the particle size corresponding to a certain cumulative volume percentage, in millimeters. The uniformity coefficient is dimensionless and determines the steepness of the curve. The larger the uniformity coefficient, the steeper the curve drops near the characteristic particle size value, indicating a more uniform block size. The smaller n is, the flatter the curve, indicating a more dispersed block size distribution. The specific values ​​of the characteristic particle size value and the uniformity coefficient are obtained by fitting using the nonlinear least squares method, serving as two characteristic parameters characterizing the blasting block size distribution.

[0052] Then extract the blasting vibration attenuation index α, such as Figure 3 As shown in the example, at least three measuring lines are arranged along the radial direction around the blasting area. A three-component vibration sensor is placed at distances of 50 meters, 100 meters, 150 meters, 200 meters, and 300 meters from the center of the blast source on each measuring line. During the blast, the peak particle velocity at each sensor is recorded. The relationship between the peak particle velocity and the proportional distance follows the Sadovsky empirical formula. Taking the natural logarithm of both sides of the formula yields a linear equation. Substituting the measured multiple sets of distances, peak particle velocities, and the known total amount of explosives in the same delay period during this blast into this linear equation, the attenuation index α can be solved through linear regression. The value of α reflects the rate of attenuation of the vibration wave in the rock mass of the mining area. The larger the α, the faster the attenuation and the smaller the vibration hazard range.

[0053] Next, the uniformity index of the charge density in the boreholes is extracted. The charge density is defined as the mass of explosive per unit length of the borehole, in kg / m. The actual charge density of each borehole is calculated based on the on-site charge records: Actual charge density = Total mass of explosive actually loaded into the borehole / Length of the charge section in the borehole. The designed charge density is obtained by dividing the theoretical charge mass of the borehole in the blasting design document by the designed charge section length. For the k-th borehole, the deviation rate is calculated as (Actual charge density - Design charge density) / Design charge density. The root mean square value of the deviation rates of all N boreholes is calculated. The uniformity index is defined as the reciprocal of the root mean square value. When the uniformity index value is large, it indicates that the actual charge density of each borehole deviates little from the design value, and the charge uniformity is good; conversely, a small uniformity index value indicates uneven charge, which may lead to uneven distribution of blasting energy.

[0054] After extracting the feature parameters, a blasting effect evaluation model is constructed to calculate the comprehensive blasting effect score. This method uses a weighted summation method to calculate the comprehensive blasting effect score. The weight coefficients are dynamically determined based on historical blasting data using the entropy weight method. First, feature parameters from the most recent 30 blasting operations are collected, including feature particle size, uniformity coefficient, attenuation index α, and charge uniformity index. Since the parameters have different dimensions, the range standardization method is used to make the original data dimensionless: for positive indicators (uniformity coefficient, attenuation index α, uniformity index), the larger the better standardization is adopted, and for negative indicators (characteristic particle size), the smaller the better standardization is adopted. Then, the entropy weight method is used to calculate the information entropy of each indicator, and then the weight coefficient of each indicator is obtained. The characteristic particle size, uniformity coefficient, attenuation index α and charge uniformity index are denoted as w1, w2, w3, and w4, respectively. The values ​​of each characteristic parameter in the current blasting operation after the same standardization process are denoted as y1', y2', y3', and y4'. The comprehensive blasting effect score S is calculated by the following formula: S = w1×y1' + w2×y2' + w3×y3' + w4×y4'. The value of S ranges from 0 to 1. The closer S is to 1, the better the blasting effect.

[0055] like Figure 4As shown, taking the data of a single blast in an open-pit iron mine as an example, the blast had a total of 120 boreholes, with a designed charge of 180 kg per borehole and an actual average charge of 178.5 kg. After preprocessing, two abnormal charge data were removed, and the missing borehole depth values ​​were filled in. The characteristic particle size obtained from the processing of the blast pile image was 187 mm, the uniformity coefficient was 1.25, and the attenuation index α obtained from the fitting of vibration monitoring data was 1.68. The root mean square value of the charge deviation rate yielded a uniformity index of 13.9. Based on historical 30 blasting data, the weights were calculated using the entropy weight method: w1=0.35, w2=0.25, w3=0.20, w4=0.20. After standardization of the parameters for the current blasting: y1'=0.547, y2'=0.5625, y3'=0.60, y4'=0.59. The comprehensive score S was calculated as: S=0.35×0.547+0.25×0.5625+0.20×0.60+0.20×0.59=0.570.

[0056] The comprehensive blasting effect score is compared with a preset threshold. If the score is lower than the threshold, a blasting parameter adjustment suggestion is generated. In this embodiment, the comprehensive score is 0.570, which is lower than the preset optimization threshold of 0.70, triggering the parameter optimization decision module to generate a blasting parameter adjustment suggestion. In this method, a decision tree model is pre-trained. The input of the model is the current blasting parameters and the comprehensive blasting effect score, and the output is the parameter adjustment suggestion. The training data comes from historical blasting records. Each record includes hole spacing, row spacing, single hole charge, detonation delay time, comprehensive blasting effect score, and the actual parameter adjustment direction given by human experts. The decision tree model adopts the CART algorithm, and the splitting criterion is minimizing the Gini coefficient. It should be noted that the specific structural parameters of the model (such as a maximum depth of 4 layers, a minimum number of leaf node samples of 3, etc.) and the 5-fold cross-validation method are all set in accordance with the conventional methods in this field. After the model is trained, for the current blasting operation, the currently used hole spacing, row spacing, average charge per hole, detonation delay time, and the calculated S value are input into the model. The model traverses the decision tree nodes, determines the direction based on the conditions of each node, and finally outputs an adjustment vector at the leaf node. This vector contains the type of blasting parameter to be adjusted, the adjustment direction (increase or decrease), and the adjustment magnitude (expressed as a percentage). For example, the output result may be: hole spacing: increase by 5%; row spacing: remain unchanged; charge per hole: decrease by 3%; detonation delay time: increase by 10 milliseconds. Figure 4 After the blasting parameters were adjusted and optimized, the overall score of the next round of blasting in the example described above improved to 0.78. Those skilled in the art can construct and train this decision tree model without excessive labor, based on the above description, to achieve automatic optimization and adjustment of blasting parameters.

[0057] In addition, there is another embodiment in which, if the comprehensive blasting effect score of the blasting parameter adjustment suggestions generated by the above decision tree model is still lower than the preset threshold, then the matching recommended parameters are retrieved from the blasting parameter knowledge base as blasting parameter adjustment suggestions; wherein the blasting parameter knowledge base is preset and stores the optimal blasting parameter range under different rock mass conditions.

[0058] Finally, the blasting parameter adjustment suggestions are output to the data receiving terminal or blasting operation control system; for example, they can be displayed to blasting engineers in a visual graphical interface, including text suggestions and parameter comparison charts; or the adjusted parameters can be directly written into the control unit of the electronic detonator initiator and the charging vehicle in the blasting operation through the industrial communication protocol to achieve closed-loop automatic control.

[0059] This solution also discloses an open-pit mine blasting data analysis system for implementing the aforementioned open-pit mine blasting data analysis method, including:

[0060] The data acquisition module is used to acquire borehole parameter data, rock mass physical and mechanical parameter data, and blasting vibration monitoring data;

[0061] The data preprocessing module is used to clean, denoise, and imput the collected data;

[0062] The feature extraction module is used to extract blasting block size distribution feature values, blasting vibration attenuation index and borehole charge density uniformity index from the preprocessed data.

[0063] The evaluation model module is used to calculate the comprehensive blasting effect score based on the feature parameters;

[0064] The optimization decision-making module is used to generate suggestions for adjusting blasting parameters based on the comprehensive blasting effect score;

[0065] The output module is used to send adjustment suggestions in a visual form or in the form of control commands to the data receiving terminal and / or the execution unit of the blasting operation.

[0066] The data acquisition module includes an UAV-borne image acquisition unit for acquiring images of the blast pile, a distributed vibration sensor network unit for acquiring peak particle velocity during blasting, and a borehole measurement unit for acquiring the mass of explosives and the length of the charge section in the borehole. The optimization decision module has a built-in blasting parameter knowledge base, which stores the optimal blasting parameter range under different rock mass conditions. When the comprehensive blasting effect score of the blasting parameter adjustment suggestion is compared with a preset threshold, if the score is lower than the threshold, the matching parameters are retrieved from the knowledge base as the blasting parameter adjustment suggestion output.

[0067] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended embodiments and their equivalents.

Claims

1. A method for analyzing blasting data in open-pit mines, characterized in that, Includes the following steps: S1. Collect borehole parameter data, rock mass physical and mechanical parameter data, and blasting vibration monitoring data in the blasting area of ​​the open-pit mine. S2. Preprocess the collected data, remove outliers and fill in missing values; S3. Extract characteristic parameters representing the blasting effect from the pre-processed data and the post-blasting data. The characteristic parameters include at least the characteristic value of blasting block size distribution, the blasting vibration attenuation index, and the uniformity index of charge density in the blast hole. S4. Construct a blasting effect evaluation model based on the aforementioned feature parameters, and calculate the comprehensive blasting effect score; S5. Compare the comprehensive blasting effect score with a preset threshold. If the score is lower than the threshold, generate a blasting parameter adjustment suggestion.

2. The method for analyzing blasting data in open-pit mines according to claim 1, characterized in that, It also includes S6: outputting the blasting parameter adjustment suggestion to the data receiving terminal and / or the blasting operation control system.

3. The method for analyzing blasting data in open-pit mines according to claim 1, characterized in that, In S1, the borehole parameter data includes the coordinate position, diameter, depth, dip angle, charge structure, detonation sequence, and charge amount per borehole for each borehole; the rock mass physical and mechanical parameter data includes the compressive strength, tensile strength, elastic modulus, Poisson's ratio, and the graded values ​​of joint and fracture development of the rock; the blasting vibration monitoring data includes vibration waveforms collected at different distances from the blasting point, peak particle velocity, dominant frequency, and vibration duration.

4. The method for analyzing blasting data in open-pit mines according to claim 1, characterized in that, In step S2, the preprocessing step includes: An algorithm based on local outlier factors is used to identify and remove outliers in borehole parameter data; Missing data is filled by a weighted average of parameters of the same type from adjacent boreholes, where the weighting coefficient is determined inversely to the borehole spacing.

5. The method for analyzing blasting data in open-pit mines according to claim 1, characterized in that, In S3, the characteristic value of the blasting block size distribution The extraction method includes characteristic particle size values ​​and a uniformity coefficient characterizing the evenness of the block size distribution. Images of the blasted piles were collected, and the rock block outlines were identified using an edge detection algorithm. The volume percentage of rock blocks of each grain size was calculated, and a block size distribution curve was obtained by fitting the curve. The characteristic grain size value and uniformity coefficient were then extracted from the curve.

6. The method for analyzing blasting data in open-pit mines according to claim 1, characterized in that, In step S3, the method for extracting the blasting vibration attenuation index is as follows: Based on the peak particle velocity data collected by vibration monitoring sensors placed at different distances from the blasting point, the vibration attenuation law is fitted by linear regression method to obtain the blasting vibration attenuation index.

7. The method for analyzing blasting data in open-pit mines according to claim 1, characterized in that, In step S3, the method for calculating the uniformity index of the charge density in the borehole is as follows: Calculate the deviation rate between the actual charge density and the designed charge density for each borehole, and calculate the root mean square value of the deviation rate for all boreholes. Use the reciprocal of this root mean square value as the uniformity index.

8. The method for analyzing blasting data in open-pit mines according to claim 1, characterized in that, In step S4, the blasting effect evaluation model adopts a weighted summation method, wherein the weight coefficients of each feature parameter are dynamically determined based on historical blasting data through the entropy weight method or the analytic hierarchy process. The comprehensive blasting effect score is the sum of the standardized values ​​of each feature parameter obtained after processing and their corresponding weights.

9. The method for analyzing blasting data in open-pit mines according to claim 1, characterized in that, In step S5, the step of generating the blasting parameter adjustment suggestion includes: The current blasting parameters and the comprehensive blasting effect score are input into a pre-trained decision tree model. The decision tree model outputs the type of blasting parameters that need to be adjusted, as well as the adjustment direction and adjustment range, according to the node splitting rules.

10. A blasting data analysis system for open-pit mines, used to implement the blasting data analysis method for open-pit mines as described in any one of claims 1 to 9, comprising: The data acquisition module is used to acquire borehole parameter data, rock mass physical and mechanical parameter data, and blasting vibration monitoring data; The data preprocessing module is used to clean, denoise, and imput the collected data; The feature extraction module is used to extract feature parameters from the data, including at least the characteristic value of blasting block size distribution, the blasting vibration attenuation index, and the uniformity index of charge density in the blast hole; The evaluation model module is used to calculate the comprehensive blasting effect score based on the feature parameters; The optimization decision-making module is used to generate suggestions for adjusting blasting parameters based on the comprehensive blasting effect score; The output module is used to send adjustment suggestions in a visual form or in the form of control commands to the data receiving terminal and / or the execution unit of the blasting operation.

Images