A stope roof significant pressure prediction method based on edge detection

Abstract

The application discloses a kind of based on edge detection's stope roof significant pressure prediction method, belong to coal mining technical field, including the following steps: step 1, obtains the real-time pressure data of electro-hydraulic control system hydraulic support;Step 2, pre-process pressure data, including outlier detection and processing, missing value processing and data denoising;Step 3, by surface fitting technology, data points are fitted into continuous surface and projection, extract the spatial distribution characteristics of roof pressure;Step 4, calculate the relevant parameters of projection area;Step 5, use edge detection technology to predict significant pressure area and update parameters.The application can more accurately predict significant pressure area, reduce misjudgment and missed case, provide more reliable basis for coal mine safety production.

Description

Technical Field

[0001] This invention belongs to the field of coal mining technology, specifically relating to a method for predicting significant roof pressure in mining areas based on edge detection. Background Technology

[0002] In the coal mining industry, the accuracy of roof pressure prediction directly affects safe production and mining efficiency. In the past, the industry generally used empirical formulas and simple statistical analysis methods to predict roof pressure. These traditional methods mainly rely on the experience of mining personnel, combined with limited hydraulic support pressure data, such as calculating the average, maximum, and minimum pressure over a period of time, and then comparing them with preset empirical thresholds to infer the probability of roof pressure.

[0003] However, coal mining environments are extremely complex, and mine pressure is influenced by a combination of factors. Different mining depths result in different pressures from overlying strata; variations in ore body morphology, such as thickness and dip angle, alter pressure distribution patterns; and mining methods, such as coal mining technology and sequence, significantly impact mine pressure. Under such complex conditions, traditional empirical formulas and basic statistical analyses are insufficient to comprehensively and accurately reflect the patterns of mine pressure variation. For example, in analyzing the three-dimensional spatial distribution of mine pressure, due to a lack of in-depth data mining capabilities, traditional methods cannot accurately pinpoint areas of drastic pressure changes, nor can they analyze the interrelationships of pressure changes in different areas, thus failing to provide comprehensive and reliable data support for coal mining.

[0004] The development of advanced technologies such as big data and machine learning has brought new ideas to the field of coal mining. Some studies have attempted to use machine learning algorithms to process mine stress data. However, the coal mining environment is unique, with difficult data collection and high noise levels. Existing machine learning models are often too complex, demanding high computing resources, and difficult to deploy and apply efficiently in the mine. Furthermore, most of these methods do not fully consider the actual geological conditions and mining processes of coal mining, resulting in unsatisfactory accuracy and reliability of the prediction results.

[0005] In coal mining, roof pressure is a key factor affecting safe production. Current traditional methods for predicting roof pressure largely rely on empirical formulas and simple statistical analysis, making it difficult to accurately predict areas of significant pressure. Existing technologies mainly suffer from the following problems.

[0006] The depth of data processing and analysis is insufficient: Current technologies for processing mine pressure data mostly remain at the level of simple statistics, such as calculating averages, maximums, and minimums. There is a lack of effective data processing algorithms and models, making it impossible to fully explore the characteristics and patterns of the data. Mine pressure is affected by various factors such as depth, ore body morphology, and mining methods, and simple statistics cannot reflect these complex relationships. For example, when analyzing the three-dimensional spatial distribution of mine pressure, it is impossible to accurately locate areas of pressure variation, failing to provide comprehensive and accurate data support for mine operations.

[0007] Limitations of 2D Analysis: Existing technologies mostly employ 2D analysis methods, which cannot fully reflect the variation of roof pressure in three-dimensional space. Roof movement is a complex three-dimensional process, and it is difficult to accurately capture its key characteristics through 2D analysis alone, resulting in insufficient accuracy and reliability of prediction results.

[0008] The predictive models lack adaptability: Most existing predictive methods are based on fixed data models and parameters, and cannot be dynamically adjusted as mining progresses and geological conditions change. This makes the predictive models prone to decreased accuracy when facing complex and ever-changing mining environments.

[0009] Based on this, the present invention provides a method for predicting significant roof pressure in mining areas based on edge detection. By analyzing historical data, significant areas are identified and characteristic values ​​are calculated. Then, based on these results, the pressure areas are further predicted, thereby improving the accuracy and reliability of the prediction and meeting the needs of the coal mining industry for safe and efficient mining. Summary of the Invention

[0010] To address the aforementioned problems, this invention proposes a method for predicting significant pressure on the mine roof based on edge detection. This method involves real-time acquisition and processing of pressure data from the hydraulic supports of the electro-hydraulic control system, constructing a three-dimensional surface using surface fitting technology, and analyzing significant pressure areas through plane projection. Simultaneously, edge detection technology is used to predict whether significant pressure will occur, identifying areas of significant pressure in the mine and providing early warnings.

[0011] The technical solution of the present invention is as follows:

[0012] A method for predicting significant roof pressure in a mining area based on edge detection, characterized by the following steps:

[0013] Step 1: Obtain real-time pressure data of the hydraulic support of the electro-hydraulic control system;

[0014] Step 2: Preprocess the stress data, including outlier detection and handling, missing value handling, and data denoising;

[0015] Step 3: Fit the data points into a continuous curved surface using surface fitting technology and project it to extract the spatial distribution characteristics of the top plate pressure.

[0016] Step 4: Calculate the relevant parameters of the projection area;

[0017] Step 5: Use edge detection technology to predict and update the significant pressure area.

[0018] Furthermore, in step 1, the pressure data of the hydraulic support is collected in real time using the pressure sensor equipped in the downhole electro-hydraulic control system at a set high-frequency sampling frequency g; within a certain advance interval, the i-th support y perpendicular to the advance direction is... i The pressure data sequence collected by the upper pressure sensor is denoted as z = [z i1 ,z i2 ,…,z ih ], where h is the number of sampling points within the advance interval, z ih Indicates the advance as x h At that time, the i-th support y i The pressure value is measured; finally, the collected pressure data is transmitted in real time and stored in the database via a data transmission network.

[0019] Furthermore, the specific process of step 2 is as follows:

[0020] Step 2.1, Outlier Detection and Handling;

[0021] For a stress data series, calculate the mean and standard deviation:

[0022]

[0023] Where, μ i Let z be the average pressure data collected by the pressure sensor on the i-th support within a certain time period; ij Indicates the advance as x j At that time, the i-th support y i The pressure value on the surface; j is the index value of the advance; σ i N represents the standard deviation of the pressure data collected by the pressure sensor on the i-th support within a certain time period; i The total number of pressure data points collected by the pressure sensor on the i-th bracket;

[0024] If |z ij -μ i |>3σ i Then determine z ij If it is an outlier, replace it with the median M of the sequence. i M i The median of all data values ​​from the pressure sensor on the i-th bracket;

[0025] Step 2.2, handling missing values;

[0026] For data points with missing values, linear interpolation is used for imputation; the advance corresponding to the missing value is defined as x. k If z ik Missing, and z i(k+1) z i(k-1) Given that, then:

[0027]

[0028] Among them, z ik z i(k+1) z i(k-1) The advance length is x. k x k+1 x k-1 At that time, the i-th support y i The pressure value; k, k+1, and k-1 are the index values ​​for different advances;

[0029] Step 2.3: Data denoising;

[0030] The moving average filtering method is used to denoise the cleaned data; assuming the moving window size is m, the denoising process is performed on each pressure value using the following formula:

[0031]

[0032] Where, z′ ij Indicates the advance as x j At that time, the i-th support y i The pressure value after denoising; z il Indicates the advance as x l At that time, the i-th support y i The pressure value; l is the index value of the advance;

[0033] When l<1 or l>N i When using forward padding, i.e., z il =z i1 or z i1 , Indicate the advance lengths are x1, ... At that time, the i-th support y i The pressure value on top.

[0034] Furthermore, the specific process of step 3 is as follows:

[0035] Step 3.1: Fit the distribution of pressure data in three-dimensional space using surface fitting technology to construct a top plate to press the curved surface;

[0036] Data points are extracted from the real-time pressure data of the hydraulic support after preprocessing in step 2. A top plate pressure surface is constructed in three-dimensional space, with the direction of advancement along the working surface set as the x-axis, the perpendicular direction of advancement set as the y-axis, and the pressure value set as the z-axis. If the advance is x... j When the i-th bracket y is collected i The pressure value is z ij Then, mapping it to three-dimensional coordinates is (x j ,y i ,z ij This allows us to construct a set of data points for the top plate to press against the curved surface.

[0037] {(x j ,y i ,z ij )|i=1,2,...,α;j=1,2,...,β};

[0038] Where α is the total number of supports; β is the total number of advance indices;

[0039] The top plate presses against the curved surface Z = f(x,y), and the fitted model is:

[0040] Z = a0 + a1x + a2y + a3x 2 +a4y 2 +a5xy;

[0041] Where f(x,y) is the pressure value collected by the pressure sensor when the support y advances by x; a0, a1, a2, a3, a4, and a5 are different coefficients to be determined;

[0042] Error function E for the construction of the curved surface under pressure on the top plate:

[0043]

[0044] Take the partial derivatives of the error function E of the top plate pressing surface with respect to the coefficients a0, a1, a2, a3, a4, and a5, and set the partial derivatives to 0 to obtain a system of six equations. Solve this system of equations to determine the values ​​of the coefficients a0, a1, a2, a3, a4, and a5, and then determine the specific expression of the top plate pressing surface Z = f(x,y).

[0045] Step 3.2, Projection Processing;

[0046] Set the pressure data value used for cutting the curved surface to D0; use a plane The top plate is cut to press the curved surface. The cut curve is projected onto a cutting plane parallel to the xoy plane to obtain the projection area. At the same time, the data point information in the projection area is recorded, including the pressure value, the corresponding support position, and the acquisition time.

[0047] Furthermore, the specific process of step 4 is as follows:

[0048] Step 4.1: Calculate the area of ​​the projected region and the mean value of the data within the region;

[0049] Divide each projection region into M×N grids, each grid having a size of Δx×Δy. Count the number K of grids falling within the projection region, then the area S of the r-th projection region is... r for:

[0050] S r = K×(Δx×Δy);

[0051] Where Δx and Δy are the length and width of each grid, respectively; M and N are the number of grids along the x-axis and y-axis, respectively;

[0052] The total area S of the projected region is the sum of the areas of all projected regions:

[0053]

[0054] Where q is the number of projection regions, and r = 1, 2, ..., q;

[0055] Calculate the average area of ​​all projected regions.

[0056]

[0057] The set of pressure data points in the r-th projection region for:

[0058]

[0059] Among them, P rg is the value of the g-th pressure data point in the r-th projection region; g is the number of pressure data points in the r-th projection region;

[0060] For the r-th projection region, the average pressure data for:

[0061]

[0062] Among them, P ro This represents the value of the o-th pressure data point in the r-th projection region;

[0063] Calculate the mean of data for all projected regions.

[0064]

[0065] Step 4.2: Comparison and Adjustment;

[0066] Calculate the total area S of the mined working face. total， The formula is:

[0067] S total =l push ×l ten ;

[0068] Among them, l push The historical advance distance of the working face; l ten The working face dip length;

[0069] Let S and S total Compare and determine if the 0.15S condition is met. total <S<0.2S total If the condition is not met, the value of D0 needs to be readjusted, and the surface cutting and projection operations need to be performed again until the condition is met. At this point, the data analysis of the historical advancement distance is complete, and the average value of the data calculated from all projected areas is recorded. and area mean

[0070] Furthermore, the specific process of step 5 is as follows:

[0071] Step 5.1: Assume the current advance in the forecasting phase is... This represents the advance index value for the prediction phase; connect the mine pressure data points for the current prediction phase into a line segment. The online database contains multiple stress data points. For line segments The first One pressure data point; It is a line segment The total number of pressure data points on the surface;

[0072] Step 5.2: Traverse the line segments using the sliding window method. Identify the significant pressure zone line segment in the prediction and endpoint positions;

[0073] Step 5.3: Determine if the current condition is met. When satisfied Then, perform the calculation in step 5.4; if Then calculate the next advance, i.e. update the advance index value. Switch to the next line segment Continue with step 5.2, performing sliding window traversal and conditional checks;

[0074] Step 5.4: Calculate the area ratio and set an early warning threshold. When the area ratio is greater than the set early warning threshold, trigger the early warning mechanism and design an early warning classification mechanism for graded early warning.

[0075] Step 5.5, if the advance is... If the area exceeds 30m, the area and data of the significant pressure zone in the predicted portion will be added to the historical data, and the mean area and mean data of the significant pressure zone will be recalculated and updated, i.e., updated once. and If advance If the advance length does not exceed 30m, update the advance index value. Switch to the next line segment After the switch is complete, for the new line segment The sliding window method from step 5.2 is used again for analysis.

[0076] Furthermore, the specific process of step 5.2 is as follows:

[0077] Step 5.2.1: Set a fixed-length sliding window with length w and step size S. Starting from the beginning of the line segment, slide the window sequentially, moving a step size 's' each time; for each window, count the number of data points exceeding a certain threshold. Number of data points c:

[0078]

[0079] Where Count(·) represents statistical calculation; Represents line segment The first One pressure data point,

[0080] Step 5.2.2, if satisfied Then record the window and the data points contained within it; if Then calculate the next advance, i.e. update the advance index value. Switch to the next line segment Execute step 5.2.1 to continue the sliding window traversal and condition judgment;

[0081] Step 5.2.3: Merge all windows that meet the conditions. The endpoints of the windows that meet the conditions are the edges of the predicted significant pressure region. The line segment connecting the endpoints is the line segment of the predicted significant pressure region. Record the endpoint positions.

[0082] Furthermore, the specific process of step 5.4 is as follows:

[0083] Calculate the advance as Area of ​​the significant pressure region predicted in time And sum them up to obtain the predicted total area S of the significant pressure region. pre,total ;definition is the warning threshold, and the area ratio is calculated as When occurs, it is determined that a large roof weighting area will appear, and a warning signal is sent in a timely manner; if occurs, it is determined that there will be no large roof weighting area at present, and no warning is given. At this time, the footage index value is switched to the next line segment and starts again from the starting point of the line segment, and step 5.2.1 is executed to traverse and conditionally judge using the sliding window method;

[0084] When predicting significant roof weighting, calculate the comprehensive prediction index, and design a warning classification mechanism based on the comprehensive prediction index;

[0085] The formula for the comprehensive prediction index is:

[0086]

[0087] where F is the comprehensive prediction index; n represents the number of step distances of the predicted significant roof weighting area; is the dynamic weight coefficient based on the relationship between the pressure data points and The calculation formula is:

[0088]

[0089] The warning classification mechanism is designed as follows:

[0090] Set the first threshold F1 and the second threshold F2 of the comprehensive prediction index. If f1 ≤ F < F2 is satisfied, it is a first-level warning, indicating local pressure concentration and requiring an increased monitoring frequency; if F ≥ F2 is satisfied, it is a second-level warning, requiring immediate stop of mining and activation of the emergency plan; if F < F1 is satisfied, no warning is given.

[0091] The beneficial technical effects brought by the present invention are as follows.

[0092] High accuracy: By collecting the pressure data of hydraulic supports in real time and using various data processing and analysis means, such as data standardization processing, surface fitting and projection, etc., the characteristics and laws of roof pressure changes can be mined. Compared with traditional methods, it can more accurately predict significant roof weighting areas, reduce misjudgment and missed judgment situations, and provide a more reliable basis for coal mine safety production.

[0093] Dynamic update: During the prediction process, when the footage exceeds a certain distance, the data of the new significant roof weighting area can be accumulated into the historical data, and relevant characteristic values are recalculated to achieve dynamic update of the prediction model. This enables the prediction method to better adapt to the changing geological conditions and mining environment during coal mine mining, and always maintain a high prediction accuracy.

[0094] Timely risk warning: The application of edge detection technology can quickly detect abrupt changes and edge features in data, identifying potential areas of concentrated pressure in advance. Once a region with potentially high pressure is predicted, a timely warning is issued, giving coal mine workers sufficient time to take effective preventative measures and reduce the risk of accidents. Attached Figure Description

[0095] Figure 1 This is a flowchart of the method for predicting significant pressure on the mining roof based on edge detection, as described in this invention.

[0096] Figure 2 The top plate pressure surface diagram is fitted to the data points of this invention.

[0097] Figure 3 This is a projection area diagram of the present invention.

[0098] Figure 4 This is a schematic diagram illustrating the overall division of the coal mining face according to the present invention.

[0099] Figure 5 This is a flowchart illustrating the significant pressure region predicted using edge detection technology in this invention.

[0100] Figure 6 This is a schematic diagram of the calculation results of the present invention as the working face advances.

[0101] Figure 7 This is a two-dimensional cloud map of the significant mining pressure region in time period 1 of Experiment 1 of this invention.

[0102] Figure 8 This is a schematic diagram of the actual pressure concentration area for some data in time period 1 of Experiment 1 of the present invention.

[0103] Figure 9 This is a schematic diagram of the predicted pressure concentration area for some data in time period 1 obtained by the edge detection method in Experiment 1 of this invention.

[0104] Figure 10 This is a two-dimensional cloud map of the significant mining pressure area in time period 2 of Experiment 2 of this invention.

[0105] Figure 11 This is a schematic diagram of the actual pressure concentration area for some data in time period 2 of Experiment 2 of this invention.

[0106] Figure 12 This is a schematic diagram of the predicted pressure concentration area for some data in time period 2 obtained by the edge detection method in Experiment 2 of this invention. Detailed Implementation

[0107] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments:

[0108] This invention introduces edge detection technology into the field of roof pressure prediction in mining areas, combining it with big data acquisition and processing technology, surface fitting technology, and other techniques to form a complete prediction system. This multi-technology integration approach breaks through the limitations of traditional prediction methods, providing a completely new approach and methodology for roof pressure prediction. This invention also establishes a dynamically updated prediction model that can continuously incorporate new data into the model for recalculation and analysis as mining progresses. This dynamic update mechanism makes the prediction model adaptive, enabling it to reflect changes in roof pressure in a timely manner and maintain the accuracy of its predictions of pressure-prone areas—a feature rarely seen in previous roof pressure prediction technologies.

[0109] This invention is a data-driven method used to predict the distribution and changing trends of future roof pressure based on existing historical data and real-time collected data. Its core lies in using mathematical and statistical methods to mine patterns and characteristics within the data, thereby enabling prediction of unknown situations. For example... Figure 1 As shown, a method for predicting significant roof pressure in a mining area based on edge detection includes the following steps:

[0110] Step 1: Obtain real-time pressure data of the hydraulic support in the electro-hydraulic control system; the specific process is as follows:

[0111] Using the high-precision pressure sensor equipped in the downhole electro-hydraulic control system, pressure data of the hydraulic support is acquired in real time according to a set high-frequency sampling frequency f. Within the advance interval [x1, x2], the i-th support y perpendicular to the advance direction is... i The pressure data sequence collected by the upper pressure sensor is denoted as z = [z i1 ,z i2 ,…,z ih ], where h is the number of sampling points within the advance interval, z ih The advance length is x h At that time, the i-th support y i The pressure value is measured; finally, the collected pressure data is transmitted in real time and stored in the database via a data transmission network.

[0112] Step 2: Preprocess the stress data, including outlier detection and handling, missing value handling, and data denoising; the specific process is as follows:

[0113] Step 2.1, Outlier Detection and Handling;

[0114] For a stress data series, calculate its mean and standard deviation:

[0115]

[0116] Where, μ iLet z be the average pressure data collected by the pressure sensor on the i-th support within a certain time period; ij Indicates the advance as x j At that time, the i-th support y i The pressure value on the surface; j is the index value of the advance; σ i Let N be the standard deviation of the pressure data collected by the pressure sensor on the i-th support within a certain time period, used to measure the fluctuation or dispersion of these data around the mean; i The total number of pressure data points collected by the pressure sensor on the i-th bracket;

[0117] If |z ij -μ i |>3σ i Then determine z ij If it is an outlier, replace it with the median M of the sequence. i M i The median of all data values ​​from the pressure sensor on the i-th bracket;

[0118] Step 2.2, handling missing values;

[0119] For data points with missing values, linear interpolation is used for imputation. The advance corresponding to a missing value is defined as x. k If z ik Missing, and z i(k+1) z i(k-1) Given that, then:

[0120]

[0121] Among them, z ik z i(k+1) z i(k-1) The advance length is x. k x k+1 x k-1 At that time, the i-th support y i The pressure value; k, k+1, and k-1 are the index values ​​for different advances;

[0122] Step 2.3: Data denoising;

[0123] The moving average filtering method is used to denoise the cleaned data. Let the size of the moving window be m (i.e., the number of observation points involved in the calculation within each advance window). The denoising process for each pressure value is performed using the following formula:

[0124]

[0125] Where, z′ ij Indicates the advance as x j At that time, the i-th support yi The pressure value after denoising; z il Indicates the advance as x l At that time, the i-th support y i The pressure value; l is the index value of the advance;

[0126] When l<1 or l>N i When using forward padding, i.e., z il =z i1 or z i1 , Indicate the advance lengths are x1, ... At that time, the i-th support y i The pressure value on top.

[0127] Step 3: Fit the data points into a continuous surface using surface fitting technology and project it to extract the spatial distribution characteristics of the top plate pressure; the specific process is as follows:

[0128] Step 3.1: Fit the distribution of pressure data in three-dimensional space using surface fitting technology to construct a top plate to press the curved surface;

[0129] Data points are extracted from the real-time hydraulic support pressure data collected by the electro-hydraulic control system after preprocessing in step 2. A top plate pressing surface is constructed in three-dimensional space. The top plate pressing surface fitted by the data points is as follows: Figure 2 As shown. The x-axis is set along the working face advancing direction (representing the advance distance along the working face, in meters), the y-axis is set perpendicular to the advancing direction (representing the support number perpendicular to the advancing direction, used to identify different hydraulic supports, in units of supports), and the z-axis is set, in MPa. Figure 2 The black area in the image represents the top plate pressing against the curved surface. If the advance is x... j When the i-th bracket y is collected i The pressure value is z ij Then, mapping it to three-dimensional coordinates is (x j ,y i ,z ij This allows us to construct a set of data points for the top plate to press against the curved surface.

[0130] {(x j ,y i ,z ij )|i=1,2,...,α;j=1,2,...,β};

[0131] Where α is the total number of supports; β is the total number of advance indices;

[0132] The top plate presses against the curved surface Z = f(x,y), and its fitting model is:

[0133] Z = a0 + a1x + a2y + a3x 2 +a4y 2 +a5xy;

[0134] Where f(x,y) is the pressure value collected by the pressure sensor when the support y is advanced by x, and the unit is megapascal (MPa); a0, a1, a2, a3, a4, and a5 are different coefficients to be determined.

[0135] Error function E for the construction of the curved surface under pressure on the top plate:

[0136]

[0137] Taking the partial derivatives of the error function E of the top plate pressing surface with respect to the coefficients a0, a1, a2, a3, a4, and a5, and setting the partial derivatives to 0, yields a system of six equations. Solving this system of equations determines the values ​​of the coefficients a0, a1, a2, a3, a4, and a5, and thus the specific expression for the top plate pressing surface Z = f(x,y).

[0138] Step 3.2, Projection Processing;

[0139] Set the pressure data value used for cutting the curved surface to D0; use a plane The predetermined top plate is cut to press the curved surface. The resulting curve is projected onto a cutting plane parallel to the xoy plane to obtain the projection area. Simultaneously, data point information within the projection area is recorded, including pressure values, corresponding support positions, and acquisition times.

[0140] Step 4: Calculate the relevant parameters of the projection area; the specific process is as follows:

[0141] Step 4.1: Calculate the area of ​​the projected region and the mean value of the data within the region;

[0142] Figure 3 This is a projection area diagram; numbers 1-20 in the diagram represent the plane. Different projection regions of the cleaving plane obtained by cutting the curved surface are divided into M×N grids, each grid having a size of Δx×Δy. The number K of grids falling within the projection region is counted, and then the area S of the r-th projection region is calculated. r for:

[0143] S r = K×(Δx×Δy);

[0144] Where Δx and Δy are the length and width of each grid, respectively; M and N are the number of grids along the x-axis and y-axis, respectively;

[0145] The total area S of the projected region is the sum of the areas of all projected regions:

[0146]

[0147] Where q is the number of projection regions, and r = 1, 2, ..., q;

[0148] Average area of ​​all projected regions It is the average area of ​​all projected regions (calculated when multiple projected regions exist; if there is only one projected region, then...). ):

[0149]

[0150] The set of pressure data points in the r-th projection region for:

[0151]

[0152] Among them, P rg It is the value of the g-th pressure data point in the r-th projection region; g is the number of pressure data points in the r-th projection region.

[0153] For the r-th projection region, the average pressure data for:

[0154]

[0155] Among them, P ro This represents the value of the o-th pressure data point in the r-th projection region;

[0156] Average data for all projected areas It is the average value of pressure data points across all projected areas:

[0157]

[0158] Step 4.2: Comparison and Adjustment;

[0159] Figure 4 This diagram illustrates the overall division of the coal mining face. Mining is carried out along the direction of the face's advance. Area A represents the historical data area, i.e., the mined area of ​​the face. Area B represents the area to be predicted, i.e., the area to be mined. All subsequent predictions will be analyzed within area B. x1, x2, and x3 represent different advances.

[0160] Calculate the total area S of the mined working face. total The formula is:

[0161] S total =l push ×l ten ;

[0162] Among them, lpush The historical advance distance of the working face; l ten The working face dip length;

[0163] Let S and S total Compare and determine if the 0.15S condition is met. total <S<0.2S total If the condition is not met, the value of D0 needs to be readjusted, and the surface cutting and projection operations need to be performed again until the condition is met. At this point, the data analysis of the historical advancement distance is complete, and the average value of the data calculated from all projected areas is recorded. and area mean

[0164] Step 5: Prediction and Parameter Update of Significant Pressure Areas. This step utilizes edge detection technology to quickly detect abrupt changes in pressure data and identify potential pressure concentration areas (significant areas). As mining progresses, new data is continuously collected and analyzed. When the advance exceeds a certain distance, new significant pressure area data is added to the historical data, and relevant feature values ​​are recalculated to achieve dynamic updates to the prediction model.

[0165] like Figure 5 This flowchart describes a method for predicting significant pressure areas using edge detection technology. It calculates the mean data and area of ​​significant areas through historical data analysis, and uses this to predict areas where large pressure may occur. Furthermore, it updates the feature values ​​of significant pressure areas as the advance reaches a certain distance. The specific process is as follows:

[0166] Step 5.1: Based on the partial step size that has already appeared, predict whether a large pressure will occur;

[0167] Assuming the current forecast phase advance is This represents the advance index value during the prediction phase; to better illustrate the progress status of the working face, the mine pressure data points for the current prediction phase are connected into line segments. The online database contains multiple stress data points. For line segments The first One pressure data point; It is a line segment The total number of pressure data points on the surface;

[0168] Step 5.2: Traverse the line segments using the sliding window method. Identify the significant pressure zone line segment in the prediction and endpoint positions; the specific process is as follows:

[0169] Step 5.2.1: Set a fixed-length sliding window with a length of w and a step size of s. ， from Starting from the beginning of the line segment, slide the window sequentially, moving a step size of 's' each time. For each window, count the number of data points exceeding a certain threshold. Number of data points c:

[0170]

[0171] Where Count(·) represents statistical calculation; Represents line segment The first One pressure data point,

[0172] Step 5.2.2, if it exceeds The number of data points c accounts for more than 80% of the total number of data points, that is, it satisfies Then record the window and the data points contained within it; if Then calculate the next advance, i.e. update the advance index value. Switch to the next line segment Execute step 5.2.1 to continue the sliding window traversal and condition judgment;

[0173] Step 5.2.3: Merge all windows that meet the conditions. The endpoints of the windows that meet the conditions are the edges of the predicted significant pressure region. The line segment connecting the endpoints is the line segment of the predicted significant pressure region. Record the endpoint positions.

[0174] Step 5.3: Determine if the current condition is met. When satisfied Then, perform the calculation in step 5.4, that is, only when... In other words, the subsequent calculation of the predicted area is only performed after at least two advance segments have been analyzed. If this condition is not met... Then calculate the next advance, i.e. update the advance index value. Switch to the next line segment Execute step 5.2.1 to continue the sliding window traversal and condition judgment;

[0175] Step 5.4: Calculate the area ratio and set an early warning threshold. When the area ratio exceeds the set threshold, an early warning mechanism is triggered. A tiered early warning mechanism is designed for tiered alerts. The specific process is as follows:

[0176] Calculate the advance as Area of ​​the significant pressure region predicted in time And sum them up to obtain the predicted total area S of the significant pressure region. pre,total ;definition is the warning threshold, and the area ratio is calculated as When occurs, it is determined that there will be a large pressure area, and a warning signal is sent in time. If occurs, it indicates that the current stage has not yet reached the condition for the possible occurrence of a large pressure area; at this time, the footage index value is updated and switched to the next line segment and start again from the starting point of the line segment, and execute step 5.2.1 to traverse and make conditional judgments using the sliding window method.

[0177] For example, Figure 6 is the schematic diagram of the calculation results as the working face advances. Figure 6 In, L salien,1 is the line segment of the predicted significant pressure area when the footage is x1; S pre,2 is the area of the predicted significant pressure area when the footage is x2; S pre,total is the total area of the predicted significant pressure area accumulated when the footage is x3.

[0178] When predicting significant pressure, factors such as the distribution of mine pressure data points and the relationship between the predicted area and historical data are comprehensively considered to calculate the comprehensive prediction index, and then the significant degree of the predicted pressure is measured; the formula for the comprehensive prediction index is:

[0179]

[0180] Among them, F is the comprehensive prediction index; n represents the number of step distances of the predicted significant pressure area (the step distance is the distance between adjacent footages). is the dynamic weight coefficient based on the relationship between the pressure data points and , and the calculation formula is:

[0181]

[0182] reflects the proportion of the pressure data exceeding the mean value at the current step distance. The higher the proportion, the greater the weight of the predicted significant pressure area at this step distance.

[0183] Warning classification mechanism:

[0184] Set the first threshold F1 and the second threshold F2 of the comprehensive prediction index. On the basis of the original warning, if F1≤F<F2 is satisfied, it is upgraded to a first-level warning, indicating local pressure concentration, and the monitoring frequency needs to be increased; if F≥F2 is satisfied, it is upgraded to a second-level warning, requiring immediate stop of mining and activation of the emergency plan; if F<F1 is satisfied, there is no warning.

[0185] Step 5.5. If the footage If the area exceeds 30m, the area and data of the significant pressure zone in the predicted portion will be added to the historical data, and the mean area and mean data of the significant pressure zone will be recalculated and updated, i.e., updated once. and If advance If the advance distance is less than 30m, it means the current data does not yet meet the conditions for updating historical data. In this case, update the advance index value. Switch to the next line segment After the switch is complete, for the new line segment The sliding window method from step 5.2 is used again for analysis.

[0186] To verify the effectiveness and reliability of this invention, mine pressure data from different time periods at the XV1311 working face were used for experiments. Experimental results show that the edge detection prediction method of this invention can successfully predict the pressure concentration areas and trends that correspond to the actual situation. Specific experimental data and analysis are as follows:

[0187] Experiment 1: Verification of mine pressure data for time period 1;

[0188] By using surface fitting and projection analysis, several significant mining pressure areas were identified from the mining pressure data of cuts 1558 (1247.0m) to 1605 (1284.6m) on the XV1311 working face, between [2025-2-3 9:23:08] and [2025-2-8 1:03:39]. Figure 7 Only a portion of the data from time period 1 is shown, revealing significant areas of two-dimensional mining pressure. The area and mean values ​​for each region were also calculated, as shown in Table 1.

[0189] Table 1 shows the data results for significant mining pressure areas in part of the data for time period 1.

[0190]

[0191]

[0192] Using the edge detection method of the present invention, pressure concentration areas are predicted based on a portion of the advance during time period 1. Figure 8 , Figure 9 The data for time period 1 shows the actual and predicted pressure concentration areas. Figure 8 and Figure 9 It can be seen that the method of the present invention predicts the pressure concentration area that matches the actual situation.

[0193] Experiment 2: Verification of mine pressure data for time period 2;

[0194] By using surface fitting and projection analysis, the mining pressure data from cuts 1604 (1283.8m) to 1678 (1343.0m) on the XV1311 working face, from [2025-2-8 0:1:54] to [2025-2-14 23:12:14], identified several significant mining pressure areas. Figure 10 The table shows the significant areas of mineral pressure in two dimensions for a portion of the data in time period 2. The area and mean values ​​of each region were also calculated, as shown in Table 2.

[0195] Table 2 shows the data results for areas of significant mine stress in part of the data in time period 2.

[0196]

[0197] Using the edge detection method of the present invention, pressure concentration areas are predicted based on a portion of the advance during time period 2. Figure 11 , Figure 12 The data for time period 2 shows the actual and predicted areas of concentrated pressure. Figure 11 and Figure 12 It can be seen that the method of the present invention predicts the pressure concentration area that matches the actual situation.

[0198] Of course, the above description is not intended to limit the present invention, and the present invention is not limited to the examples given above. Any changes, modifications, additions or substitutions made by those skilled in the art within the scope of the present invention should also fall within the protection scope of the present invention.

Claims

1. A method for predicting significant roof pressure in a mining area based on edge detection, characterized in that, Includes the following steps: Step 1: Obtain real-time pressure data of the hydraulic support of the electro-hydraulic control system; Step 2: Preprocess the stress data, including outlier detection and handling, missing value handling, and data denoising; Step 3: Fit the data points into a continuous curved surface using surface fitting technology and project it to extract the spatial distribution characteristics of the top plate pressure. The specific process is as follows: Step 3.1: Fit the distribution of pressure data in three-dimensional space using surface fitting technology to construct a top plate to press the curved surface; Data points are extracted from the real-time pressure data of the hydraulic support after preprocessing in step 2. A top plate pressure surface is constructed in three-dimensional space, with the direction of advancement along the working surface set as the x-axis, the perpendicular direction of advancement set as the y-axis, and the pressure value set as the z-axis. If the advance is... The first time to collect One bracket pressure value Then, mapping it to three-dimensional coordinates is This allows us to construct a set of data points for the top plate to press against the curved surface: ； in, Total number of stents; This represents the total number of advance indices; This is the index value for the advance. Top plate to press curved surface The fitted model is: ； in, For stent In advance The pressure value collected by the pressure sensor; , , , , , For different coefficients to be determined; Error function for pressing curved surface construction on top plate : ； Error function of the pressure surface on the top plate Regarding coefficients , , , , , Find the partial derivatives and set them to zero to obtain a system of six equations; solve the system of equations to determine the coefficients. , , , , , The value of is used to determine the pressure of the top plate on the curved surface. The specific expression; Step 3.2, Projection Processing; Set the pressure data value used for cutting the curved surface to be [value]. ; using a plane The top plate is cut to press the curved surface. The resulting curve is projected onto a cutting plane parallel to the xoy plane to obtain the projection area. At the same time, the data point information within the projection area is recorded, including the pressure value, the corresponding support position, and the acquisition time. Step 4: Calculate the relevant parameters of the projection area; the specific process is as follows: Step 4.1: Calculate the area of ​​the projected region and the mean value of the data within the region; Divide each projection area into There are 1 grid, each grid having a size of 1. Count the number of grids falling within the projection area. Then the first Area of ​​each projection area for: ； in, , These represent the length and width of each grid cell, respectively. , These represent the number of grid cells along the x-axis and y-axis, respectively. Total area of ​​the projected region It is the sum of the areas of all projected regions: ； in, The number of projection areas. ; Calculate the average area of ​​all projected regions. : ； No. A set of pressure data points in a projection area for: ； in, It is the first The first projection region The value of each pressure data point; It is the first Number of pressure data points in each projection area; For the Average pressure data for each projected area for: ； in, For the first The first projection region The value of each pressure data point; Calculate the mean of data for all projected regions. : ； Step 4.2: Comparison and Adjustment; Calculate the total area of ​​the mined working face. The formula is: ； in, This represents the historical advance distance of the working face. The working face dip length; Will and Compare and determine whether the conditions are met. If this condition is not met, it needs to be readjusted. The value is then used to perform surface cutting and projection operations again until the conditions are met. At this point, the data analysis of the historical advancement distance is complete, and the average value of the data calculated for all projected areas is recorded. and area mean ; Step 5: Utilize edge detection technology to predict and update salient pressure regions; the specific process is as follows: Step 5.1: Assume the current advance in the forecasting phase is... , This represents the advance index value for the prediction phase; connect the mine pressure data points for the current prediction phase into a line segment. The online system contains multiple stress data points. ; For line segments The first One pressure data point; It is a line segment The total number of pressure data points on the surface; Step 5.2: Traverse the line segments using the sliding window method. Determine the significant pressure area line segment in the prediction. and endpoint positions; Step 5.3: Determine if the current condition is met. When satisfied Then, perform the calculation in step 5.4; if Then calculate the next advance, i.e., update the advance index value. Switch to the next line segment Continue with step 5.2, performing sliding window traversal and conditional checks; Step 5.4: Calculate the area ratio and set an early warning threshold. When the area ratio is greater than the set early warning threshold, trigger the early warning mechanism and design an early warning classification mechanism for graded early warning. Step 5.5, if the advance is... If the area exceeds 30m, the area and data of the significant pressure zone in the predicted portion will be added to the historical data, and the mean area and mean data of the significant pressure zone will be recalculated and updated, i.e., updated once. and If the advance is made... If the advance length does not exceed 30m, update the advance index value. Switch to the next line segment After the switch is complete, for the new line segment Then, the sliding window method from step 5.2 is used again for analysis.

2. The method for predicting significant roof pressure in a mining area based on edge detection according to claim 1, characterized in that, In step 1, the pressure sensor equipped with the downhole electro-hydraulic control system is used to sample at a set high-frequency frequency. The pressure data of the hydraulic support is collected in real time; within a certain advance interval, the first [axis] perpendicular to the advance direction is [measured / implemented]. One bracket The pressure data sequence collected by the upper pressure sensor is denoted as ,in This represents the number of sampling points within the advance interval. Indicates the advance as At that time, the first One bracket The pressure value is measured; finally, the collected pressure data is transmitted in real time and stored in the database via a data transmission network.

3. The method for predicting significant roof pressure in a mining area based on edge detection according to claim 2, characterized in that, The specific process of step 2 is as follows: Step 2.1, Outlier Detection and Handling; For a stress data series, calculate the mean and standard deviation: ； ； in, For the first The average pressure data collected by the pressure sensors on each bracket within a certain time period; Indicates the advance as At that time, the first One bracket The pressure value on top; For the first The standard deviation of pressure data collected by pressure sensors on a bracket within a certain time period; For the first The total number of pressure data points collected by the pressure sensors on each bracket; like Then determine If it is an outlier, replace it with the median of the sequence. ; For the first The median of all data values ​​from the pressure sensors on the bracket; Step 2.2, handling missing values; For data points with missing values, linear interpolation is used for imputation; the advance corresponding to the missing value is defined as... ,like Missing, and , Given that, then: ； in, , , The advance lengths are respectively , , At that time, the first One bracket The pressure value on top; , , These are the index values ​​for different advances; Step 2.3: Data denoising; The moving average filtering method is used to denoise the cleaned data; let the moving window size be... For each pressure value, noise reduction is performed using the following formula: ； in, Indicates the advance as At that time, the first One bracket The pressure value after noise reduction; Indicates the advance as At that time, the first One bracket The pressure value on top; This is the index value for the advance. when or When this is the case, a forward fill method is used, that is... or ; , The advance lengths are respectively , At that time, the first One bracket The pressure value on top.

4. The method for predicting significant roof pressure in a mining area based on edge detection according to claim 3, characterized in that, The specific process of step 5.2 is as follows: Step 5.2.1: Set a sliding window of fixed length, the window length is... The step size of the sliding window is ,from Starting from the beginning of the line segment, slide the window sequentially, moving the step size each time. For each window, the number of data points within the statistical window exceeds [a certain threshold]. Number of data points : ； in, Indicates statistical calculation; Represents line segment The first One pressure data point, ; Step 5.2.2, if satisfied If so, then record the window and the data points contained within it; if Then calculate the next advance, i.e., update the advance index value. Switch to the next line segment Execute step 5.2.1 to continue the sliding window traversal and condition judgment; Step 5.2.3: Merge all windows that meet the conditions. The endpoints of the windows that meet the conditions are the edges of the predicted significant pressure region. The line segment connecting the endpoints is the line segment of the predicted significant pressure region. Record the endpoint positions.

5. The method for predicting significant roof pressure in a mining area based on edge detection according to claim 4, characterized in that, The specific process of step 5.4 is as follows: Calculate the advance as Area of ​​the significant pressure region predicted in time And sum them up to obtain the predicted total area of ​​the significantly compressed region. ;definition The area ratio is calculated as the warning threshold. ;when If this occurs, it is determined that a large area of ​​pressure will appear, and a warning signal will be issued in a timely manner; if If no large pressure area is detected, no warning will be issued, and the advance index value will be updated. Switch to the next line segment , from Starting from the beginning of the line segment, execute step 5.2.1, using the sliding window method for traversal and conditional judgment; When significant pressure is predicted, a comprehensive forecasting index is calculated, and an early warning grading mechanism is designed based on the comprehensive forecasting index. The formula for calculating the comprehensive forecast index is as follows: ； in, As a comprehensive forecasting indicator; This indicates the number of steps in the predicted significant pressure area; Based on pressure data points and The dynamic weighting coefficient of the relationship is calculated using the following formula: ； The early warning grading mechanism is designed as follows: Set the first threshold for the comprehensive forecasting index Second threshold If satisfied If the conditions are met, it will be a Level 1 warning, indicating concentrated local pressure and requiring increased monitoring frequency; if the conditions are met... If the conditions are met, it is a Level II warning, requiring an immediate halt to mining and activation of the emergency response plan; if the conditions are met... In such cases, no warning will be issued.

Images