A mine roof stability evaluation method, device and equipment

Abstract

The present application relates to the field of data processing, and provide a kind of mine roof stability evaluation method, device and equipment, the method comprises: obtaining the multiple monitoring indexes of mine roof pressure, determines original index according to monitoring index;Monitoring index includes step sequence resistance, step sequence average resistance, safety valve opening pressure, roof deformation, impact acceleration and roof beam offset angle;The original index is normalized, and standardization data is obtained;Standardization data is input into trained neural network model, and output stability prediction result;Stability prediction result is scored based on objective function to obtain score result, and roof stability grade is determined according to score result;Objective function is designed according to the optimal range distribution law of multiple monitoring indexes.The present application solves the problem of low efficiency of manual evaluation of roof stability in the prior art, realizes intelligent dynamic collaborative analysis and intelligent grading early warning based on multi-source monitoring index.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a method, apparatus and equipment for evaluating the stability of a mine roof. Background Technology

[0002] Roof stability management is a core aspect of coal mine safety production, directly impacting the lives of underground workers and mining efficiency. During coal mining, dynamic changes in roof pressure can lead to disasters such as rock strata movement, sudden changes in support load, and even roof collapse. Traditional roof pressure monitoring primarily relies on the working resistance data of hydraulic supports, but a single parameter is insufficient to comprehensively reflect the dynamic behavior of the roof, especially under complex conditions such as periodic pressure inrushes and rock bursts, where the monitoring accuracy and early warning capabilities of existing technologies are significantly inadequate.

[0003] Currently, pressure monitoring of mine roofs mainly employs physical similarity simulation experiments and digital monitoring systems. Physical similarity simulation experiments use scaled-down models to simulate the mining process, observing rock fractures and support stress. However, this approach faces three major technical bottlenecks: it relies on manual observation of rock fissures or intermittent data acquisition, failing to capture transient processes of roof instability (such as impact acceleration and dynamic opening of safety valves); the determination of roof pressure intensity is largely based on empirical indicators (such as fracture morphology), lacking quantitative analysis models and resulting in poor repeatability; and the boundary conditions and load transfer mechanisms of the similarity simulation differ from actual mining, leading to low reliability of monitoring results. While some digital monitoring systems have incorporated pressure sensors or displacement measurement devices, their functionality is limited, only acquiring static parameters (such as end-step resistance), failing to integrate multi-dimensional dynamic indicators (such as deformation velocity and impact acceleration), and lacking comprehensive analysis and intelligent early warning capabilities.

[0004] Therefore, how to achieve simultaneous monitoring of multiple indicators of roof pressure, dynamic quantitative evaluation, and intelligent graded early warning has become a core requirement for the industry to break through the limitations of traditional technologies. Summary of the Invention

[0005] This invention provides a method, apparatus, and equipment for evaluating the stability of mine roofs, which solves the problem of low efficiency in manual evaluation of roof stability in the prior art, and realizes intelligent dynamic collaborative analysis and intelligent hierarchical early warning based on multi-source monitoring indicators.

[0006] This invention provides a method for evaluating the stability of a mine roof, comprising the following steps:

[0007] Multiple monitoring indicators of mine roof pressure are obtained, and raw indicators are determined based on the monitoring indicators; the monitoring indicators include end resistance of step sequence, average resistance of step sequence, safety valve opening pressure, roof deformation, impact acceleration, and roof beam offset angle.

[0008] The original indicators are normalized to obtain standardized data;

[0009] The standardized data is input into a trained neural network model, which outputs a stability prediction result. The neural network model is trained based on a training set, which includes standardized data and stability prediction labels corresponding to multiple monitoring indicators.

[0010] The stability prediction results are scored based on the objective function to obtain a score result, and the roof stability level is determined based on the score result; the objective function is designed based on the optimal range distribution law of multiple monitoring indicators.

[0011] According to the present invention, a method for evaluating the stability of a mine roof slab includes training a neural network model based on a training set. Specifically, this includes: constructing an initial neural network model; determining the consistency constraints of the neural network model based on the convergence time of each monitoring indicator; inputting the standardized data into the initial neural network model, training it based on the consistency constraints, and outputting predicted values; obtaining the difference between the predicted values ​​and the back supervision function value; if the difference is less than a preset threshold, temporarily stopping training and saving the trained neural network model; the back supervision function is used to force the neural network to follow physical laws during training and prediction through the consistency constraints; performing an objective function test on the predicted values ​​output by the neural network model; iteratively training the neural network model and performing an objective function test based on the neural network model obtained in each iteration; when the number of iterations reaches a preset number, ending the iteration loop and saving the trained neural network model.

[0012] According to the present invention, a method for evaluating the stability of a mine roof is provided, wherein the monitoring indicators of the roof are directly collected by sensors; the original indicators include the periodic pressure step distance, dynamic load coefficient, roof surface fracture degree, roof deformation rate, resistance rise rate at the end of the step sequence, pressure rise rate before the safety valve opens, roof beam offset angular velocity, roof impact strength, safety valve opening time ratio, and non-uniform roof pressure.

[0013] According to the method for evaluating the stability of a mine roof provided by the present invention, the determination of the original indicators based on the monitoring indicators specifically includes: determining the periodic pressure step distance based on the step-end resistance or the step-end average resistance; determining the dynamic load coefficient based on the step-end average resistance during the periodic pressure period and the non-pressure period of the roof; determining the degree of roof surface fragmentation based on the roof deformation; determining the roof deformation rate based on the roof deformation; determining the step-end resistance rise rate based on the step-end resistance; determining the pressure rise rate before the safety valve opens based on the safety valve opening pressure; determining the roof beam offset angular velocity based on the roof beam offset angle; determining the roof impact intensity based on the impact acceleration; determining the average safety valve opening time ratio based on the safety valve opening pressure; and determining the non-uniform roof pressure based on the roof beam offset angle.

[0014] According to a method for evaluating the stability of a mine roof provided by the present invention, the step of determining the periodic pressure step distance based on the step-end resistance or the step-average resistance specifically includes: obtaining the average value and range of the step-average resistance; determining the main discrimination index for periodic pressure based on the average value and range of the step-average resistance; when the step-average resistance is greater than the main discrimination index for periodic pressure, determining that periodic pressure has occurred on the roof, and obtaining the periodic pressure step distance; or obtaining the average value and range of the step-end resistance; determining the auxiliary discrimination index for periodic pressure based on the average value and range of the step-end resistance; when the step-end resistance is greater than the auxiliary discrimination index for periodic pressure, determining that periodic pressure has occurred on the roof, and obtaining the periodic pressure step distance.

[0015] According to the method for evaluating the stability of a mine roof provided by the present invention, the step of determining the dynamic load coefficient based on the step-by-step average resistance during the period of periodic roof pressure and the period of non-pressure of the roof specifically includes: obtaining the ratio of the step-by-step average resistance during the period of periodic roof pressure and the period of non-pressure of the roof as the dynamic load coefficient.

[0016] According to the present invention, a method for evaluating the stability of a mine roof slab, wherein determining the degree of roof surface breakage based on the roof deformation specifically includes: obtaining the range and average value of the roof deformation; and determining the ratio of the range and average value of the roof deformation as the degree of roof surface breakage.

[0017] According to a method for evaluating the stability of a mine roof provided by the present invention, when obtaining the stability prediction label, the method further includes: when the periodic pressure step distance is not less than a first pressure step distance threshold, determining that the roof stability belongs to level one stability; when the periodic pressure step distance is less than the first pressure step distance threshold but greater than a second pressure step distance threshold, determining that the roof stability belongs to level two stability; when the periodic pressure step distance is not greater than the second pressure step distance threshold, determining that the roof stability belongs to unstable.

[0018] According to a method for evaluating the stability of a mine roof provided by the present invention, when obtaining the stability prediction label, the method further includes: when the dynamic load coefficient is not greater than a first dynamic load coefficient threshold, determining that the roof stability belongs to level one stability; when the dynamic load coefficient is greater than the first dynamic load coefficient threshold and less than a second dynamic load coefficient threshold, determining that the roof stability belongs to level two stability; and when the dynamic load coefficient is not less than the second dynamic load coefficient threshold, determining that the roof stability belongs to unstable.

[0019] According to a method for evaluating the stability of a mine roof provided by the present invention, when obtaining the stability prediction label, the method further includes: when the degree of surface breakage of the roof is not greater than a first surface breakage threshold, determining that the roof stability belongs to Level 1 stability; when the degree of surface breakage of the roof is greater than the first surface breakage threshold and less than a second surface breakage threshold, determining that the roof stability belongs to Level 2 stability; and when the degree of surface breakage of the roof is not less than the second surface breakage threshold, determining that the roof stability belongs to instability.

[0020] According to a method for evaluating the stability of a mine roof provided by the present invention, when obtaining the stability prediction label, the method further includes: determining a first relative error between the adjacent time displacement difference method and the displacement fitting curve differentiation method based on the roof deformation velocity; determining a first relative deviation between the maximum value and the average value of the roof deformation velocity based on the value of the first relative error; determining that the roof stability belongs to level one stability when the first relative deviation is not greater than a first relative deviation threshold; determining that the roof stability belongs to level two stability when the first relative deviation is greater than the first relative deviation threshold and less than a second relative deviation threshold; and determining that the roof stability belongs to unstable stability when the first relative deviation is not less than the second relative deviation threshold.

[0021] According to a method for evaluating the stability of a mine roof provided by the present invention, when obtaining the stability prediction label, the method further includes: determining a second relative error between the adjacent time displacement difference method and the displacement fitting curve differentiation method based on the rate of increase of resistance at the end of the step; determining a second relative deviation between the maximum value and the average value of the rate of increase of resistance at the end of the step based on the value of the second relative error; determining that the roof stability is level one when the second relative deviation is not greater than a third relative deviation threshold; determining that the roof stability is level two when the second relative deviation is greater than the third relative deviation threshold and less than a fourth relative deviation threshold; and determining that the roof stability is unstable when the second relative deviation is not less than the fourth relative deviation threshold.

[0022] According to a method for evaluating the stability of a mine roof provided by the present invention, when obtaining the stability prediction label, the method further includes: determining the third relative error between the adjacent time displacement difference method and the displacement fitting curve differentiation method based on the pressure rise rate before the safety valve opens; determining the third relative deviation between the maximum value and the average value of the pressure rise rate before the safety valve opens based on the value of the third relative error; determining that the roof stability is level one when the third relative deviation is not greater than the fifth relative deviation threshold; determining that the roof stability is level two when the third relative deviation is greater than the fifth relative deviation threshold and less than the sixth relative deviation threshold; and determining that the roof stability is unstable when the third relative deviation is not less than the sixth relative deviation threshold.

[0023] According to a method for evaluating the stability of a mine roof provided by the present invention, when obtaining the stability prediction label, the method further includes: when the roof impact intensity is not greater than a first impact intensity threshold, determining that the roof stability belongs to level one stability; when the roof impact intensity is greater than the first impact intensity threshold and less than a second impact intensity threshold, determining that the roof stability belongs to level two stability; and when the roof impact intensity is not less than the second impact intensity threshold, determining that the roof stability belongs to unstable.

[0024] According to a method for evaluating the stability of a mine roof provided by the present invention, when obtaining the stability prediction label, the method further includes: obtaining the average value of the safety valve opening time ratio; when the average value of the safety valve opening time ratio is not greater than a first time ratio average value threshold, determining that the roof stability belongs to Level 1 stability; when the average value of the safety valve opening time ratio is greater than the first time ratio average value threshold and less than a second time ratio average value threshold, determining that the roof stability belongs to Level 2 stability; when the average value of the safety valve opening time ratio is not less than the second time ratio average value threshold, determining that the roof stability belongs to unstable.

[0025] According to the present invention, a method for evaluating the stability of a mine roof is provided. The step of determining the roof stability level based on the scoring results specifically includes: the stability levels are Level 1, Level 2, and Level 3; if the objective function score is not less than a first threshold and all consistency constraints of the neural network model are satisfied, the roof stability level is determined to be Level 1; if the objective function score is not less than a second threshold and all consistency constraints of the neural network model are satisfied, the roof stability level is determined to be Level 2; if the objective function score is less than a third threshold or any consistency constraint of the neural network model is violated, the roof stability level is determined to be Level 3.

[0026] The present invention also provides a device for evaluating the stability of a mine roof, comprising the following modules:

[0027] The indicator acquisition module is used to acquire multiple monitoring indicators of the pressure on the top of the mine roof and determine the original indicators based on the monitoring indicators. The monitoring indicators include the end resistance of the step sequence, the average resistance of the step sequence, the opening pressure of the safety valve, the amount of roof deformation, the impact acceleration, and the offset angle of the roof beam.

[0028] The standardization module is used to normalize the original indicators to obtain standardized data;

[0029] The model prediction module is used to input the standardized data into a trained neural network model and output a stability prediction result; the neural network model is trained based on a training set, which includes standardized data and stability prediction labels corresponding to multiple monitoring indicators.

[0030] The performance evaluation module is used to score the stability prediction results based on the objective function to obtain a score result, and to determine the roof stability level based on the score result; the objective function is designed based on the optimal range distribution law of multiple monitoring indicators.

[0031] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the mine top stability evaluation method as described above.

[0032] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the mine roof stability evaluation method as described above.

[0033] The present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the mine top stability evaluation method as described above.

[0034] This invention provides a method, apparatus, and equipment for evaluating the stability of mine roofs, which offers the following advantages: By comprehensively acquiring multi-dimensional monitoring indicators such as end-of-step resistance, average step-of-step resistance, safety valve opening pressure, roof deformation, impact acceleration, and roof beam offset angle, and performing normalization processing and neural network modeling analysis, a multi-parameter collaborative intelligent evaluation of mine roof stability is achieved. This overcomes the limitations of traditional single-indicator evaluations. Utilizing a trained neural network model and objective function scoring mechanism, it can more comprehensively and objectively reflect the dynamic changes in the roof, ultimately outputting scientific graded early warning results. This significantly improves the accuracy and reliability of roof stability evaluation, reduces human analysis costs, and increases performance prediction efficiency, providing intelligent decision support for safe mining. Attached Figure Description

[0035] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0036] Figure 1 This is a schematic diagram of the model architecture of the mine roof pressure monitoring system provided by the present invention.

[0037] Figure 2 This is a schematic diagram of the data acquisition device provided by the present invention.

[0038] Figure 3 This is a schematic diagram of the power system structure of the pump station provided by the present invention.

[0039] Figure 4 This is a partial structural diagram of the power system of the pump station provided by the present invention.

[0040] Figure 5 This is a schematic diagram of the model support provided by the present invention.

[0041] Figure 6 This is a schematic diagram of the software monitoring interface and real-time data collection list provided by the present invention.

[0042] Figure 7 This is a schematic diagram of the parameter calibration of the model support provided by the present invention.

[0043] Figure 8 This is a schematic diagram of the parameter calibration and analysis of the model support provided by the present invention.

[0044] Figure 9 This is a flowchart illustrating the method for evaluating the stability of a mine roof provided by this invention.

[0045] Figure 10 The step-end resistance p provided by this invention m —t relationship curve.

[0046] Figure 11 This is a schematic diagram illustrating the principle of solving the upward velocity of the resistance at the end of the step provided by the present invention.

[0047] Figure 12 This is a p-t relationship curve diagram of two different sequences provided by the present invention.

[0048] Figure 13 This is a schematic diagram illustrating the principle of solving the pressure rise rate before the safety valve opens, provided by the present invention.

[0049] Figure 14 This is a Pt curve diagram before and after the safety valve is opened, provided by the present invention.

[0050] Figure 15 This is the Sn-t relationship curve of the top plate deformation provided by the present invention.

[0051] Figure 16 This is a schematic diagram illustrating the principle of solving the descent speed of the column under contraction provided by the present invention.

[0052] Figure 17 This is the impact acceleration a-t relationship curve provided by the present invention.

[0053] Figure 18 This is a statistical curve of the top beam offset angle at any time provided by the present invention.

[0054] Figure 19 This is a statistical curve of the pressure on the model support provided by the present invention.

[0055] Figure 20 This is a schematic diagram illustrating the principle of solving the top beam offset angular velocity provided by the present invention.

[0056] Figure 21 This is a flowchart of the algorithm for evaluating the performance of mine roof slabs provided by the present invention.

[0057] Figure 22 This is a schematic diagram of the optimal performance range of the model support provided by the present invention.

[0058] Figure 23 This is a schematic diagram of the structure of the mine roof stability evaluation device provided by the present invention.

[0059] Figure 24 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

[0060] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0061] Roof management is a crucial element of coal mine safety, and its importance cannot be ignored, as it directly relates to worker safety. At coal mine sites, hydraulic supports, with their built-in hydraulic cylinders and systems, provide powerful support, effectively holding up the mine roof structure and preventing roof collapse accidents. The use of physical model experiments to deeply explore key areas such as the movement patterns of overlying strata, the interaction mechanism between supports and surrounding rock, the pressure resistance of the roof, and the rationality of support strength has been widely validated as an efficient and scientific research method.

[0062] However, existing mining similarity model experiments still face the following technical bottlenecks that urgently need to be addressed: ① Traditional simulation experiments lack real-time monitoring methods for roof stability, making it difficult to accurately capture the dynamic process of roof pressure and its instability, and thus failing to meet the needs of current experiments; ② The determination of roof pressure intensity lacks a standardized system for quantitative evaluation. Existing methods mostly rely on observing the morphology of rock fractures and surface fissures, which suffers from strong subjectivity and poor repeatability, resulting in insufficient reliability of experimental data; ③ There are discrepancies between the physical model and the actual working conditions, and the reliability of existing roof pressure monitoring systems in the study of mine pressure manifestation patterns is low.

[0063] To address the aforementioned technical challenges, this invention innovatively develops a mine roof pressure monitoring system based on multi-index monitoring, and constructs a multi-dimensional evaluation method for roof stability based on the acquired monitoring indicators. This monitoring system employs a novel model architecture, enabling simultaneous acquisition of working resistance, safety valve opening status, roof deformation, and impact acceleration. It analyzes and judges different indicator categories, and combines a comprehensive evaluation method to construct an evaluation method containing multiple quantitative indicators, achieving dynamic hierarchical early warning of roof safety status.

[0064] The following is combined with Figure 1 for Figure 22 The embodiments of the present invention are described in detail.

[0065] 1.1 Model Composition

[0066] The model architecture of the mine roof pressure monitoring system consists of four parts: a power system, a control device, a hydraulic support, and a evacuation device. Figure 1 As shown.

[0067] The hydraulic support model 23 is placed at the simulated mining face. The pump station power system 21 is connected to the hydraulic control device 22 to provide power to the model device. The hydraulic control device 22 is connected to the support model 23 and can control the extension and retraction of the support columns. The evacuation device 24 is connected to the control device 22 and can quickly drain water from the model device. Figure 2 As shown.

[0068] (1) Power system

[0069] The power system of the pumping station mainly includes pumps, pressure stabilizers, and on / off valves (V0, V3, V6, V9, V12). The inlet provides water as the pressure transmission medium for the entire system. The pressurized water from the pumps is connected to the hydraulic control device via a conduit. The pumps provide the required water pressure to the hydraulic system, providing sufficient power for the lifting and lowering of the model support, and also allowing the setting of the overflow pressure (Pa, Pb, Pc, Pd) of the back pressure valve. The pressure stabilizer maintains a constant output pressure through an accumulator. During long-term support operations, the pumping station's power system may experience pressure fluctuations due to water leakage, temperature changes, or slow load variations. The pressure stabilizer dynamically adjusts to automatically replenish or release pressure, ensuring that the support force of the supports on the roof remains at the set value, guaranteeing the support effect. Simultaneously, the pressure stabilizer can coordinate the synchronous operation of multiple supports and improve system response speed, such as... Figure 3 As shown.

[0070] (2) Control device

[0071] The control device includes a five-way valve, a back pressure valve, on / off valves (V1, V4, V7, V10), and pressure sensors (P1, P2, P3, P4). The first end of the five-way valve is connected to the booster pump via the on / off valves (V1, V4, V7, V10), providing power water pressure to the entire hydraulic control device. The second end of the five-way valve is connected to the lower end of the back pressure valve, transmitting pressure from the hydraulic control device to the back pressure valve. The third end of the five-way valve is connected to the model support, controlling its lifting and lowering. The fourth end of the five-way valve is connected to the pressure sensors (P1, P2, P3, P4), accurately monitoring the pressure inside the model support columns. The fifth end of the five-way valve is connected to a vacuum device, which, when activated, quickly empties the water from the entire hydraulic control device. The back pressure valve has a pressure of Pa (or Pb, Pc, Pd) at one end and P1 (or P2, P3, P4) at the other end. When P1 > Pa, the back pressure valve opens and the safety valve overflows. Figure 4 As shown.

[0072] (3) Evacuation device

[0073] The evacuation device includes a vacuum pump, a water tank, and switching valves (V2, V5, V8, V11). One port of the water tank is connected to one port of a five-way valve via the switching valves (V2, V5, V8, V11). When the valves are opened, the hydraulic control device is connected to the evacuation device. The other port of the water tank is connected to the vacuum pump. When the vacuum pump starts, air is extracted from the water tank, creating negative pressure, and water from the hydraulic control device flows into the water tank through a conduit. Opening the switching valves allows the water in the hydraulic control device to be drained, enabling the rapid descent of the model support beam and the complete retraction of the columns. Figure 3 As shown.

[0074] (4) Hydraulic support

[0075] The hydraulic support model mainly includes a top beam 1, two hydraulic cylinders 2, a water pipe 3, a base 4, and a protective device 7. The bottom ends of the cylinder bodies of the two hydraulic cylinders 2 are fixed to the base 4 at intervals, with their center lines parallel to the long side of the base and located in the center. The top ends of the piston rods of the hydraulic cylinders are connected and fixed to the top beam 1. The two hydraulic cylinders 2 are connected through the water pipe 3; the water pipe 3 is connected to the aforementioned hydraulic control device. The protective device 7 is fixed to the rear of the top beam with screws, providing protection for the data lines of the monitoring system below, preventing large rocks from collapsing and damaging the cables during the experiment. Figure 5 As shown.

[0076] 1.2 Monitoring System

[0077] The model support condition monitoring device includes a pressure sensor 5, a laser rangefinder 6, and an acceleration sensor 8, such as... Figure 5 As shown.

[0078] One pressure sensor P is connected to water pipe 3 to measure the pressure of the hydraulic cylinder of the model support, while another pressure sensor P is connected to the other end of the back pressure valve to measure the pressure of the back pressure valve in the control device. Figure 4 As shown, the pressure sensor readings can be processed by calculation software to provide the corresponding pressure.

[0079] The laser rangefinder 6 is mounted on the base 4, positioned on the center line of the two hydraulic cylinders, directly behind the cylinders, and perpendicular to the base to ensure the laser beam is aligned with the top beam. When the hydraulic cylinders drive the top beam to move, the laser rangefinder continuously emits laser light at a high frequency and receives reflected signals, allowing it to measure the displacement of the top beam and obtain dynamic data on the extension of the front and rear columns.

[0080] Accelerometer 8 is fixed at the geometric center of the side of top beam 1 to ensure that the measurement point is located on the axis of structural symmetry, avoiding measurement deviations caused by asymmetrical loads or local deformations. It also ensures that the sensor is rigidly connected to the top beam to avoid signal distortion due to loosening. The sensor's sensitive axis is strictly aligned with the vertical direction to monitor the instantaneous acceleration changes of the top beam in the vertical plane.

[0081] 1.3 Signal Processing System

[0082] The signal processing system uses computer software to monitor and process data from pressure sensors, laser rangefinders, and attitude detectors in real time. The signal outputs of these sensors are connected to a computer via cables. This computer has a data storage device and a control interface. The data storage device stores the monitored and calculated data, and the control interface controls the operating status of the hydraulic control device. Figure 6 As shown.

[0083] 2.1 Parameter Calibration

[0084] The operating parameters of the model support were calibrated using a calibration device, which included a force sensor reaction frame, height adjustment pads, and other components. During calibration, the water pressure in the pump station was adjusted to change the water pressure on the columns, gradually increasing the water pressure from 0 MPa. The pressure sensor recorded the water pressure on the columns, and the force sensor recorded the working resistance of the support. To verify the stability of the model support performance, each model support was calibrated repeatedly using the same method at least three times. Figure 7 As shown.

[0085] The repeatability calibration results should demonstrate that the test data exhibits good stability, and that the mechanical properties of the model support are stable and meet experimental requirements. The time history curve of the support with "column water pressure as output load" is shown below. Figure 8 As shown in the figure, the hydraulic pressure of the column and the output load of the support maintain a good linear relationship, and the linear regression correlation coefficient is required to reach above 0.95.

[0086] 2.2 Operational Steps for Obtaining Monitoring Indicators

[0087] A similar simulation experiment was conducted using four model supports as a group to illustrate the experimental acquisition operations for each monitoring indicator. The overall structure is as follows: Figure 3 As shown.

[0088] 1. After the working face of the physical similarity simulation experiment is cut, the hydraulic support model is placed into the simulated mining face, the computer software is turned on, the signal monitoring system is started, and the equipment data is cleared.

[0089] 2. Open the water inlet and wait until the power pump is completely filled with clean water, then immediately close the water inlet. Next, open the main switch valve V0 to prepare for the following operations;

[0090] 3. Set the opening pressure P of the safety valve. k Open the switch valves V3, V6, V9, and V 12 The pressure is increased by a power pump until the readings of pressure sensors Pa, Pb, Pc, and Pd stabilize at the preset pressure value P. k ;

[0091] 4. Extend the support columns of the control model outwards. During this process, valves V3, V6, V9, and V6 must be closed. 12 And open valves V1, V4, V7, and V 10 The power pump was used to pressurize the system again, ensuring the model support could be raised smoothly.

[0092] 5. Set the initial support force of the model support to P0. After the top beam of the model support is in complete contact with the overlying rock layer, observe the rate of increase of the readings of pressure sensors P1, P2, P3, and P4. The rate of increase will slow down as the initial support force P0 is approached. Once the initial support force P0 is reached and stabilized, immediately close valves V1, V4, V7, and V...10 The pressure on the support is locked by physical means to maintain the stability of the model support and avoid interference from factors such as accidental touch.

[0093] 6. After the excavation face is excavated and the model support stabilizes, or if the roof collapses, the safety valve of the model support will open to obtain a stable end-step resistance P. m The vacuum device is activated, which causes the columns of the No. 1 and No. 3 model supports to descend rapidly and advance to the predetermined position. Adjustments are made to ensure that the model supports are accurately positioned.

[0094] 7. Pressurize the model supports again using a power pump to enable the No. 1 and No. 3 top beams to quickly and effectively support the rock strata above the newly excavated working face, ensuring that the required initial support force P0 is achieved. Following this, the No. 2 and No. 4 model supports are sequentially lowered, moved, and raised to continue the entire support operation process.

[0095] 8. Wait for a period of time to observe the working status of the model support. After obtaining the displacement of the model support column, continue to excavate the working face forward. The model support will repeat the operation process of steps 6 and 7. Multiple experiments are conducted to avoid accidental errors.

[0096] Figure 9 This is a flowchart illustrating the mine roof stability evaluation method provided by the present invention, as shown below. Figure 9 As shown, the method includes the following steps:

[0097] S910. Obtain multiple monitoring indicators of mine roof pressure and determine the original indicators based on the monitoring indicators.

[0098] The monitoring indicators include the end resistance of the step sequence, the average resistance of the step sequence, the opening pressure of the safety valve, the deformation of the top plate, the impact acceleration, and the offset angle of the top beam.

[0099] Specifically, in order to comprehensively assess the pressure intensity of the mine roof, it is necessary to use a variety of monitoring indicators for analysis.

[0100] 3.1 End-of-step resistance P m

[0101] (1) Magnitude of resistance at the end of the step sequence

[0102] End-of-step resistance, a key indicator of working resistance, specifically refers to the resistance state before the support is moved at the successful completion of each excavation step. Under normal operating conditions, this resistance value represents the maximum working resistance level within a single step, and its data can be collected and recorded using precision pressure sensors. Real-time tracking and monitoring of end-of-step resistance provides an intuitive and effective basis for assessing the pressure intensity of the roof above the model support.

[0103] Once the model support is in close contact with the roof slab and reaches the preset initial support force p0, a collapse of the roof slab will exert a strong impact on the top beam of the model support, causing the support columns to drop instantaneously and accompanied by a significant increase in the support force. Figure 10 As shown. Through real-time measurement by a pressure sensor, the end-of-step resistance P can be accurately captured. m Size.

[0104] When multiple model supports exist in a similar model test, the final resistance generated by the model supports at different positions differs when subjected to the impact of the top plate in the same step sequence. The average value of the final resistance is calculated, and the maximum value P is selected. m,max With minimum value P m,min And calculate the range.

[0105] Average resistance at the end of the step ;

[0106] Extreme difference in resistance at the end of the step ;

[0107] In the formula, , … The step-end resistance of the model support at different locations is expressed in megapascals (MPa); n represents the total number of model supports used in the similar model test.

[0108] (2) The rate of increase of resistance at the end of the step

[0109] The model support rises from the set initial support force p0 to p m During the process, the resistance at the end of the step exhibits an upward velocity. Different top plate pressure intensities will produce different rates of increase in the final resistance of the sequence. A faster rate of increase indicates a higher top plate pressure intensity, while a slower rate indicates a lower rate of increase. The rate of increase in the final resistance of the sequence is used to measure the top plate pressure intensity, and the maximum value P of the final resistance of the sequence is selected. m,max The model support was studied.

[0110] Calculate the average rate of increase in resistance at the end of the step sequence. Based on this, it is also necessary to solve for the velocity of the resistance at the end of the step sequence at any time, and find its maximum value. Based on the formula for calculating the velocity at any moment when the resistance rises at the end of the step sequence, the standard deviation of the velocity at the end of the step sequence is obtained. ,like Figure 11 As shown.

[0111] Average speed of rise of resistance at the end of the step ;

[0112] At any moment, the resistance rises at the end of the step sequence, increasing the velocity. ;

[0113] Standard deviation of the rate of ascent at the end of the step resistance ;

[0114] In the formula, , Record when the resistance at the end of the step does not reach the set value. The time interval between adjacent moments, in seconds (S); , When the resistance at the end of the step does not reach the set value, the pressure sensor records... The working resistance corresponding to adjacent time points, in MPa; , … The velocity at any moment during the ascent phase when the resistance at the end of the step does not reach the set value, expressed in megapascals per second (MPa·S). -1； n represents the number of data points involved in the calculation; Δp′ represents the change in resistance; and Δt represents the change in time.

[0115] 3.2 Step-by-step average resistance P t

[0116] (1) Magnitude of average resistance in sequence

[0117] During the operation of the model support, the resistance to the support changes continuously over time, and the final resistance at the end of each step is insufficient to fully reflect the stress state of the model support. The final resistance of the same model support in two different steps may be similar or equal, but the stress state of the support within the step may differ significantly, such as... Figure 12 As shown. Step-by-step average resistance. This can reflect the difference.

[0118] Step-wise average resistance refers to the average working resistance over a single step, calculated using time as the unit. P can be calculated based on the relationship between working resistance and time. t Its value is the area enclosed by the curve divided by the total time the force is applied. To simplify the calculation, the area enclosed by the curve is divided into several trapezoids, and the result is obtained using the formula:

[0119] Step average resistance ;

[0120] In the formula, , … The duration corresponding to each curved trapezoid, in seconds; , … The resistance of the model support at different times is expressed in MPa.

[0121] (2) Average value of step-by-step average resistance

[0122] When multiple model supports exist in a similar model test, the average resistance generated by the model supports at different positions will vary when subjected to the impact of the top plate in the same step sequence. The average value of the average resistance is calculated, and the average resistance P is selected. t,max With minimum value P t,min And calculate the range.

[0123] Average value of step average resistance

[0124] Range of step-by-step average resistance

[0125] In the formula, , … The step average resistance of the model support at different locations is expressed in MPa; n is the total number of model supports used in the similar model test.

[0126] 3.3 Safety valve opening pressure P k

[0127] (1) Pressure rise rate before the safety valve opens

[0128] The working resistance of the model support increases from the set initial support force P0 to the safety valve opening pressure P. k During the process, the safety valve experiences an upward velocity before reaching the opening pressure. Different pressure intensities from the top plate will result in different pressure rise rates before the safety valve opens. A faster rate indicates a higher pressure intensity from the top plate, while a slower rate indicates a lower rate. The pressure rise rate before the safety valve opens is used to measure the pressure intensity from the top plate. The average pressure rise rate before the safety valve opens is then calculated. Based on this, it is also necessary to solve for the rate of pressure rise at any given moment before the safety valve opens, and find its maximum value. Based on the formula for calculating the rate of pressure rise at any moment before the safety valve opens, the standard deviation of the rate of pressure rise before the safety valve opens is obtained. ,like Figure 13 As shown.

[0129] Average rate of pressure rise before safety valve opens ;

[0130] Speed ​​of pressure rise at any moment before safety valve opens ;

[0131] Standard deviation of pressure rise rate before safety valve opens ;

[0132] In the formula, t n-1 t n+1Record t before the safety valve opens. n The time interval between adjacent moments, in seconds (s); p n-1 p n+1 Before the safety valve opens, the pressure sensor records t. n The working resistance corresponding to adjacent time points, in MPa; , … The rate of pressure rise at any given moment before the safety valve opens, expressed in MPa·s. -1 Δp is the change in the opening pressure of the safety valve; Δt is the change in time.

[0133] (2) Average ratio of safety valve opening time

[0134] Due to the impact from the top plate, the working resistance of the model column rises from the initial support force P0 to the opening pressure P of the safety valve. k The safety valve opens. After opening, the pressure curve of the model support oscillates in a sawtooth pattern until it converges and stabilizes, at which point the next cycle begins. Figure 14 As shown.

[0135] When multiple model supports exist, the average safety valve opening time ratio η t Calculate using the following formula:

[0136]

[0137] In the formula, t1 is the pressure P required for the column's working resistance to rise from P0 to the opening pressure of the safety valve. k Record the moment of the first arrival, in seconds; t2 is the time when the working resistance of the column rises from P0 to P. k When the value drops to P0 again, the time is recorded in seconds; m is the number of model supports in the similar model test where the safety valve is not opened; n is the total number of model supports used in the similar model test.

[0138] 3.4 Roof Deformation Sn

[0139] (1) Displacement of the top plate deformation

[0140] The deformation of the top plate refers to the change in deformation of the top plate in the vertical direction. Due to the initial support force P0, the top plate and the top beam of the model support in the similar model test are in close contact. Therefore, the deformation Sn of the top plate in the similar model test can be represented by the shrinkage of the support column. This data can be collected in real time and accurately using a laser rangefinder. The deformation of the top plate is not only a key indicator for measuring the degree of column shrinkage during compression, but also one of the important bases for evaluating the compressive strength of the top plate.

[0141] By monitoring the dynamic changes in roof deformation in real time, the deformation characteristics of the roof can be intuitively understood. Furthermore, in-depth analysis of roof deformation indicators can provide a more detailed assessment of roof pressure monitoring, such as... Figure 15 As shown.

[0142] When multiple model supports exist in a similar model test, the deformation displacement of the top plate caused by the top plate impact at different positions of the model supports will also differ when subjected to the top plate pressure impact in the same step sequence. The average value of the top plate deformation is calculated, and the maximum value Sn of the top plate deformation is selected. max With minimum value Sn min And calculate the range.

[0143] Average value of top plate deformation ;

[0144] The deformation of the top plate is extremely poor. ;

[0145] In the formula, , … The deformation of the top plate of the model support at different locations is expressed in mm; n is the total number of model supports used in the similar model test.

[0146] (2) Deformation speed of the top plate V sn

[0147] When the roof collapses as a whole, the model support columns in close contact with it will respond rapidly. During the descent of the model support and the roof together, they can be considered as a single unit; at this point, the vertical deformation rate of the roof is equal to the descent rate of the model support columns. During this process, the roof deformation rate... This has become an important indicator for measuring the compressive strength and deformation of the roof slab. By calculating the descent speed of the column at any given moment, the maximum deformation rate of the roof slab can be determined. To further assess the dynamic changes in roof pressure, such as Figure 16 As shown.

[0148] Average deformation rate of top plate ;

[0149] Top plate deformation velocity at any time ;

[0150] Standard deviation of top plate deformation rate ;

[0151] In the formula, ΔSn is the change in the amount of shrinkage of the column; Δt is the change in time. , The reduction in column size did not reach the set value. At that time, the laser rangefinder recorded Time intervals between adjacent moments, in seconds (s). , The reduction in column size did not reach the set value. At that time, the laser rangefinder recorded The column descent amount at adjacent time points, in mm; , … The reduction in column size did not reach the set value. At any given moment during the contraction phase, the velocity is expressed in mm·s. -1 .

[0152] 3.5 Impact acceleration a

[0153] (1) Magnitude of impact acceleration

[0154] The intensity of the pressure from the top plate can also be measured using the impact acceleration α. ​​Due to the initial support force P0, the top plate is in close contact with the top beam of the model support, and can be considered as a single unit at the moment of impact. The impact acceleration of the top plate is obtained by an acceleration sensor placed on the top beam of the model support.

[0155] The impact acceleration was plotted as a curve. In the interval 0 to t1, the top plate experienced a strong impact, with the instantaneous acceleration gradually decreasing from its maximum value to zero, resulting in a rapid descent of the top plate. The descent velocity reached its maximum at time t1. In the interval t1 to t2, due to the stiffness of the model support, the top plate experienced acceleration in the opposite direction of the impact, slowing its descent. The descent velocity reached its minimum value of zero at time t2. Figure 17 As shown. When multiple model supports exist in a similar model test, the impact acceleration generated by the model supports at different positions will also differ when subjected to the impact of the top plate in the same step. The average impact acceleration of each model support at the same moment is calculated, and the impact acceleration 'a' is selected. max With minimum value a min And calculate the range.

[0156] Average impact acceleration ;

[0157] Impact acceleration range ;

[0158] In the formula, , … The impact acceleration of the model support at different locations at the same moment is expressed in millimeters per square second (mm·s). -2 n represents the total number of model supports used in the similar model experiment, in units of 1.

[0159] 3.6 Top beam offset angle θ

[0160] The offset angle of the top beam of the model support is defined as the angle between the centerline of the top beam and the centerline of the model support base on the horizontal plane. It is a key indicator for measuring the degree of horizontal offset of the top beam relative to the base. This data can be collected and analyzed in real time using an electronic accelerometer, converting the acceleration signal into an angle value. When the top beam tilts, the gravitational components of each axis of the accelerometer change. Using the horizontal plane as a reference, the angle is calculated based on the proportional relationship of the gravitational components. , where a x a y and a z The triaxial angular velocity values ​​are measured by the accelerometer, and the included angle θ is the offset angle of the top beam.

[0161] The offset angle θ of the top beam indirectly reflects the non-uniformity of pressure distribution through the overall deformation of the top beam, and can be used to measure the dynamic response characteristics of the roof's spatial stability. For example... Figure 18 As shown, the value of the top beam offset angle at any time in the precise step sequence is obtained. and extreme values Then, the average value of the top beam offset angle can be calculated. This provides strong support for a comprehensive assessment of roof stability.

[0162] Average value of top beam offset angle ;

[0163] In the formula, θ is the offset angle of the top beam, in degrees. ° .

[0164] According to the method for evaluating the stability of a mine roof provided by the present invention, the monitoring indicators of the roof are directly collected by sensors; the original indicators include the periodic pressure step distance, dynamic load coefficient, roof surface fracture degree, roof deformation rate, resistance rise rate at the end of the step, pressure rise rate before the safety valve opens, roof beam offset angular velocity, roof impact strength, safety valve opening time ratio, and non-uniform roof pressure.

[0165] According to the method for evaluating the stability of a mine roof provided by the present invention, the original indicators are determined based on monitoring indicators, specifically including: determining the periodic pressure step distance based on the end-step resistance or the average step resistance; determining the dynamic load coefficient based on the average step resistance during the periodic pressure period and the non-pressure period of the roof; determining the degree of roof surface fragmentation based on the roof deformation; determining the roof deformation rate based on the roof deformation; determining the rate of increase of the end-step resistance based on the end-step resistance; determining the rate of pressure increase before the safety valve opens based on the safety valve opening pressure; determining the roof beam offset angular velocity based on the roof beam offset angle; determining the roof impact intensity based on the impact acceleration; determining the average safety valve opening time ratio based on the safety valve opening pressure; and determining the non-uniform roof pressure based on the roof beam offset angle.

[0166] 4.1 Periodic pressure step distance D

[0167] According to the present invention, a method for evaluating the stability of a mine roof includes determining the periodic pressure step distance based on the step-end resistance or the step-average resistance. Specifically, the method includes: obtaining the average value and range of the step-average resistance; determining the main discrimination index for periodic pressure based on the average value and range of the step-average resistance; determining that periodic pressure has occurred when the step-average resistance is greater than the main discrimination index, and obtaining the periodic pressure step distance; or obtaining the average value and range of the step-end resistance; determining the auxiliary discrimination index for periodic pressure based on the average value and range of the step-end resistance; determining that periodic pressure has occurred when the step-end resistance is greater than the auxiliary discrimination index, and obtaining the periodic pressure step distance.

[0168] Specifically, the analysis cycle for pressure step distance uses the average value of the average resistance of each step of the support and the resistance at the end of the step as the main reference indicators. The sum of the average value of the average resistance of each step and its half-range is used as the main discrimination indicator; the sum of the average value of the resistance at the end of the step and its half-range is used as the auxiliary discrimination indicator.

[0169] Key indicators for judging cyclical pressure:

[0170]

[0171] In the formula, R is the average of the step-by-step average resistance, in MPa. pt This represents the range of the step-by-step average resistance, expressed in MPa.

[0172] Periodic pressure auxiliary judgment indicators:

[0173]

[0174] In the formula, R is the average resistance at the end of the step, in MPa. pm This represents the range of resistance at the end of the step, expressed in MPa.

[0175] The fact that the average resistance in the sequence is greater than the periodic pressure is a key indicator for determining whether periodic pressure is occurring on the roof. For example... Figure 19 As shown, it can be observed in steps 17 and 23. Based on this, the excavation distance in one step can be obtained when the roof pressure step distance D is 6 times; alternatively, an auxiliary criterion can be used when... When the resistance at the end of the step is greater than the resistance criterion at the end of the step, the top plate also experiences periodic pressure. Both methods can determine the periodic pressure step distance D.

[0176] According to the method for evaluating the stability of a mine roof provided by the present invention, when obtaining the stability prediction label, the method further includes: when the periodic pressure step distance is not less than the first pressure step distance threshold, the roof stability is determined to be level one stable; when the periodic pressure step distance is less than the first pressure step distance threshold and greater than the second pressure step distance threshold, the roof stability is determined to be level two stable; when the periodic pressure step distance is not greater than the second pressure step distance threshold, the roof stability is determined to be unstable.

[0177] Specifically, the stability standard of the roof is determined based on the pressure step distance D of the roof:

[0178] when At that time, it can be determined that the roof stability is relatively good, but the intensity of a single pressure surge may be higher, so it is necessary to focus on preventing rockbursts.

[0179] when At this time, it can be determined that the roof is relatively stable, the hazard is small, and the possibility of an accident is low;

[0180] when At that time, it can be determined that the frequency of periodic failures is higher, the roof is unstable, and more frequent reinforcement of support is required.

[0181] in and This is a critical value that can be set according to requirements. For example, the initial pressing step distance d can be used as the critical value for the fixed-cycle pressing step distance D. The initial pressing step distance d can be obtained based on the distance the working face advances during the initial pressing of the test. and The critical step size is determined by two cycles; where d is the initial step size.

[0182] 4.2 Dynamic load factor N

[0183] According to the method for evaluating the stability of a mine roof provided by the present invention, the dynamic load coefficient is determined based on the step-by-step average resistance during the period of periodic roof pressure and the period of non-pressure. Specifically, the method includes obtaining the ratio of the step-by-step average resistance during the period of periodic roof pressure and the period of non-pressure as the dynamic load coefficient.

[0184] Specifically, the ratio of the average resistance of the support sequence during the period of pressure on the basic roof to that during the period of no pressure is called the dynamic load coefficient.

[0185]

[0186] In the formula, The average resistance value of the support sequence during the periodic pressure on the roof is expressed in MPa. This represents the average resistance value of the support sequence during the non-pressurized period of the roof, in MPa.

[0187] According to the method for evaluating the stability of a mine roof provided by the present invention, when obtaining the stability prediction label, if the dynamic load coefficient is not greater than the first dynamic load coefficient threshold, the roof stability is determined to be level one stable; if the dynamic load coefficient is greater than the first dynamic load coefficient threshold and less than the second dynamic load coefficient threshold, the roof stability is determined to be level two stable; if the dynamic load coefficient is not less than the second dynamic load coefficient threshold, the roof stability is determined to be unstable.

[0188] Specifically, the dynamic load factor, which measures the ratio between the actual dynamic load and the static load, reflects the magnitude of the basic roof pressure. Based on the dynamic load factor N, different critical values ​​can be determined to assess the stability of the roof.

[0189] when When the roof is stable and poses little risk, an accident is unlikely to occur.

[0190] when At that time, it can be determined that the roof is relatively stable, the risk is small, and the possibility of an accident is low;

[0191] when If the roof is unstable, the damage is serious, and an accident is likely to occur.

[0192] in and This is a critical value that can be set according to the actual situation.

[0193] For example:

[0194] when At that time, it was determined that the roof was stable, the static load was much greater than the dynamic load, and an accident was unlikely to occur;

[0195] when When the roof is determined to be in a potentially unstable state, it is necessary to strengthen monitoring and adjust the support parameters.

[0196] when If the roof is determined to be in a high-probability unstable state, and a sudden collapse or rockburst may occur, emergency measures must be taken.

[0197] 4.3 Degree of surface breakage of the top plate ω

[0198] According to the present invention, a method for evaluating the stability of a mine roof slab is provided, which determines the degree of roof surface breakage based on the roof deformation. Specifically, the method includes: obtaining the range and average value of the roof deformation; and determining the ratio of the range and average value of the roof deformation as the degree of roof surface breakage.

[0199] Specifically, the range of deformation R of the top plate Sn The ratio of the deformation to the average deformation Sn of the top plate can be used to describe the degree of surface fragmentation of the top plate during impact. Based on the degree of surface breakage of the top plate This is used to determine the stability of the roof slab.

[0200]

[0201] According to the method for evaluating the stability of a mine roof provided by the present invention, when obtaining the stability prediction label, if the degree of surface breakage of the roof is not greater than the first surface breakage threshold, the roof stability is judged to be level one stable; if the degree of surface breakage of the roof is greater than the first surface breakage threshold but less than the second surface breakage threshold, the roof stability is judged to be level two stable; if the degree of surface breakage of the roof is not less than the second surface breakage threshold, the roof stability is judged to be unstable.

[0202] Specifically, when When the roof is stable and poses little risk, an accident is unlikely to occur.

[0203] when At that time, it can be determined that the roof is relatively stable, the risk is small, and the possibility of an accident is generally low.

[0204] when If the roof is unstable, the damage is serious, and an accident is very likely to occur.

[0205] in and This is a threshold value that can be set according to requirements.

[0206] For example:

[0207] when When the top plate is stable, the deformation distribution is relatively uniform.

[0208] when When the deformation dispersion of the roof plate increases significantly and micro-cracks begin to appear in local areas, it is necessary to adjust the support strategy.

[0209] when At that time, it can be determined that a through crack has formed on the surface of the roof, and the degree of fragmentation is extremely high. Measures should be taken to deal with the risk of instantaneous collapse or rockburst.

[0210] 4.4 Velocity Analysis

[0211] (1) Relative deviation Δ1 of the deformation rate of the top plate

[0212] According to the method for evaluating the stability of a mine roof provided by the present invention, when obtaining stability prediction labels, the first relative error between the displacement difference method and the displacement fitting curve differentiation method between adjacent time intervals is determined based on the roof deformation rate; the first relative deviation between the maximum value and the average value of the roof deformation rate is determined based on the value of the first relative error; when the first relative deviation is not greater than the first relative deviation threshold, the roof stability is judged to be of level one stability; when the first relative deviation is greater than the first relative deviation threshold and less than the second relative deviation threshold, the roof stability is judged to be of level two stability; when the first relative deviation is not less than the second relative deviation threshold, the roof stability is judged to be unstable.

[0213] Specifically, the adjacent time displacement difference method is sensitive to measurement noise and easily affected by instantaneous disturbances, which may lead to large fluctuations in the roof deformation rate. Therefore, the displacement fitting curve differentiation method is introduced, and the combination of the two improves the accuracy of roof stability determination.

[0214] Based on the obtained top plate deformation data, a drawing is performed, such as... Figure 15 As shown. Obtain the fitted curve. Required goodness of fit Differentiating the fitted curve yields the velocity at any moment of roof deformation. Calculate the maximum value of the top plate deformation rate of the fitted curve. and the average deformation rate of the top plate of the fitted curve .

[0215]

[0216] In the formula, and H1 represents the average velocity obtained by the adjacent time displacement difference method and the velocity at any time, while H1 represents the relative error between the adjacent time displacement difference method and the displacement fitting curve differentiation method.

[0217] The value of H1, used to determine the relative error, can be chosen based on actual needs, such as a critical value. .

[0218] when This indicates that the velocity curves of both methods show a high degree of consistency, meaning that the displacement data has low noise, the deformation pattern is stable, and the instantaneous fluctuations of the difference method match the trend of the fitting method well. In short-time abrupt changes, the displacement difference method with adjacent time intervals is chosen to find the maximum velocity. Because the direct method preserves the details of the original data, it can more sensitively capture the instantaneous peak value of the roof deformation, while the fitting method may underestimate the extreme values ​​due to the smoothing effect.

[0219] The stability of the roof is quantified by the relative deviation (denoted as Δ1) between the maximum and average values ​​of the roof deformation rate.

[0220]

[0221] Based on the value of parameter Δ1, the condition of the roof is divided into three levels, for example:

[0222] This indicates that the speed fluctuation is small, the deformation is uniform, and the top plate has good stability;

[0223] This indicates increased velocity dispersion, local disturbances in the top plate, and generally poor stability.

[0224] This indicates that the speed fluctuates drastically, the pressure from the top plate is intense, and the risk of instability is high.

[0225] when The significant difference in velocity curves between the two methods is likely due to data noise or deviation of the fitted model from the actual deformation pattern. Therefore, the displacement fitting curve differentiation method should be chosen to find the maximum velocity value. The fitting method, by smoothing noise and removing outliers, more reliably reflects the overall trend of roof deformation and avoids spurious extreme values ​​caused by local disturbances in the difference method, thus making it more suitable for assessing roof stability.

[0226] The stability of the roof is quantified by the relative deviation (denoted as Δ´1) between the maximum and average values ​​of the roof deformation rate.

[0227]

[0228] Based on the value of parameter Δ´1, the top plate condition is divided into three levels, for example:

[0229] This indicates that the speed fluctuation is small, the deformation is uniform, and the top plate has good stability;

[0230] This indicates increased velocity dispersion, local disturbances in the top plate, and generally poor stability.

[0231] This indicates that the speed fluctuates drastically, the pressure from the top plate is intense, and the risk of instability is high.

[0232] (2) The relative deviation Δ2 of the resistance rise speed at the end of the step sequence

[0233] According to the method for evaluating the stability of a mine roof provided by the present invention, when obtaining stability prediction labels, the second relative error between the displacement difference method and the displacement fitting curve differentiation method at adjacent times is determined based on the rate of increase of resistance at the end of the step; the second relative deviation between the maximum value and the average value of the rate of increase of resistance at the end of the step is determined based on the value of the second relative error; when the second relative deviation is not greater than the third relative deviation threshold, the roof stability is judged to be level one; when the second relative deviation is greater than the third relative deviation threshold and less than the fourth relative deviation threshold, the roof stability is judged to be level two; when the second relative deviation is not less than the fourth relative deviation threshold, the roof stability is judged to be unstable.

[0234] Specifically, the adjacent time displacement difference method is sensitive to measurement noise and easily affected by instantaneous disturbances, which may lead to large fluctuations in the resistance rise rate at the end of the step. Therefore, the displacement fitting curve differentiation method is introduced, and the combination of the two improves the accuracy of the top plate stability determination.

[0235] Plotting is performed based on the acquired step-end resistance data, such as... Figure 10 As shown. Obtain the fitted curve. Required goodness of fit Differentiating the fitted curve yields the velocity at any moment of the final resistance step. Calculate the maximum rate of increase of resistance at the end of the step in the fitted curve. The average rate of increase of resistance at the end of the fitted curve step sequence .

[0236]

[0237] in, and H1 represents the average velocity obtained by the adjacent time displacement difference method and the velocity at any given time. H2 represents the relative error between the adjacent time displacement difference method and the displacement fitting curve differentiation method. The critical value H2 for determining the relative error can be chosen according to actual needs; for example, it can be set to... .

[0238] when This indicates that the velocity curves of both methods show a high degree of consistency, meaning that the resistance data at the end of the step sequence has low noise and a stable upward trend, and the instantaneous fluctuations of the difference method match the trend of the fitting method well. In short-time abrupt changes, the adjacent time displacement difference method is chosen to find the maximum velocity. Because the direct method preserves the details of the original data, it can more sensitively capture the instantaneous peak of the resistance rise at the end of the step, while the fitting method may underestimate the extreme values ​​due to the smoothing effect;

[0239] The stability of the top plate is quantified by the relative deviation (denoted as Δ2) between the maximum and average values ​​of the resistance rise rate at the end of the step sequence.

[0240]

[0241] Based on the parameter Δ2 value, the top plate condition is divided into three levels, for example:

[0242] This indicates that the speed fluctuation is small, the deformation is uniform, and the top plate has good stability;

[0243] This indicates increased velocity dispersion, local disturbances in the top plate, and generally poor stability.

[0244] This indicates that the speed fluctuates drastically, the pressure from the top plate is intense, and the risk of instability is high.

[0245] when The significant difference in velocity curves between the two methods is likely due to data noise or deviation of the fitted model from the actual deformation pattern. Therefore, the displacement fitting curve differentiation method should be chosen to find the maximum velocity value. The fitting method, by smoothing noise and removing outliers, more reliably reflects the overall trend of roof deformation and avoids spurious extreme values ​​caused by local disturbances in the difference method, thus making it more suitable for assessing roof stability.

[0246] The stability of the top plate is quantified by the relative deviation (denoted as Δ´2) between the maximum and average values ​​of the resistance rise rate at the end of the step sequence.

[0247]

[0248] Based on the value of parameter Δ´2, the top plate condition is divided into three levels, for example:

[0249] This indicates that the speed fluctuation is small, the deformation is uniform, and the top plate has good stability;

[0250] This indicates increased velocity dispersion, local disturbances in the top plate, and generally poor stability.

[0251] This indicates that the speed fluctuates drastically, the pressure from the top plate is intense, and the risk of instability is high.

[0252] (3) The relative deviation of the pressure rise rate Δ3 before the safety valve opens

[0253] According to the method for evaluating the stability of a mine roof provided by the present invention, when obtaining the stability prediction label, the third relative error of the displacement difference method and the displacement fitting curve differentiation method between adjacent time periods is determined based on the pressure rise rate before the safety valve opens; the third relative deviation between the maximum value and the average value of the pressure rise rate before the safety valve opens is determined based on the value of the third relative error; when the third relative deviation is not greater than the fifth relative deviation threshold, the roof stability is judged to be level one stable; when the third relative deviation is greater than the fifth relative deviation threshold and less than the sixth relative deviation threshold, the roof stability is judged to be level two stable; when the third relative deviation is not less than the sixth relative deviation threshold, the roof stability is judged to be unstable.

[0254] Specifically, the adjacent time displacement difference method is sensitive to measurement noise and easily affected by instantaneous disturbances, which may lead to large fluctuations in velocity values. Therefore, the displacement fitting curve differentiation method is introduced, and the combination of the two improves the accuracy of the roof stability determination.

[0255] A graph is plotted based on the pressure data obtained before the safety valve opens, such as... Figure 13 As shown. Obtain the fitted curve. Required goodness of fit Differentiating the fitted curve yields the pressure rise rate at any moment before the safety valve opens. Calculate the maximum rate of pressure rise before the safety valve opens, based on the fitted curve. The average rate of pressure rise before the safety valve opens, and the fitted curve. .

[0256]

[0257] in, and H3 represents the average velocity obtained by the adjacent time displacement difference method and the velocity at any given time. H3 is the relative error between the adjacent time displacement difference method and the displacement fitting curve differentiation method. The critical value H3 for determining the relative error can be chosen according to actual needs; for example, it can be set to... .

[0258] when This indicates that the velocity curves of both methods show a high degree of consistency, meaning that the pressure data before the safety valve opens has low noise and a stable upward trend, and the instantaneous fluctuations of the difference method match the trend of the fitting method well. In short-term abrupt changes, the adjacent time displacement difference method is chosen to find the maximum velocity. Because the direct method preserves the details of the original data, it can more sensitively capture the instantaneous peak value of the pressure rise before the safety valve opens, while the fitting method may underestimate the extreme values ​​due to the smoothing effect;

[0259] The stability of the roof is quantified by the relative deviation (denoted as Δ3) between the maximum and average values ​​of the pressure rise rate before the safety valve opens.

[0260]

[0261] Based on the parameter Δ3 value, the condition of the roof is divided into three levels, for example:

[0262] This indicates that the speed fluctuation is small, the deformation is uniform, and the top plate has good stability;

[0263] This indicates increased velocity dispersion, local disturbances in the top plate, and generally poor stability.

[0264] This indicates that the speed fluctuates drastically, the pressure from the top plate is intense, and the risk of instability is high.

[0265] when The significant difference in the velocity curves between the two methods is likely due to data noise or deviation of the fitted model from the actual deformation pattern. Therefore, the method of finding the maximum velocity value using the derivative of the fitted curve is chosen. The fitting method, by smoothing noise and removing outliers, more reliably reflects the overall trend of roof deformation and avoids spurious extreme values ​​caused by local disturbances in the difference method, thus making it more suitable for assessing roof stability.

[0266] The stability of the roof is quantified by the relative deviation (denoted as Δ´3) between the maximum and average values ​​of the pressure rise rate before the safety valve opens.

[0267]

[0268] Based on the value of parameter Δ´3, the condition of the roof is divided into three levels, for example:

[0269] This indicates that the speed fluctuation is small, the deformation is uniform, and the top plate has good stability;

[0270] This indicates increased velocity dispersion, local disturbances in the top plate, and generally poor stability.

[0271] This indicates that the speed fluctuates drastically, the pressure from the top plate is intense, and the risk of instability is high.

[0272] (4) The relative deviation Δ4 of the top beam offset angular velocity

[0273] The offset angle of the top beam at any time value It can calculate the angular velocity of the top beam offset angle at any given time. This allows us to find the maximum value of the top beam's offset angular velocity. To further understand the dynamic stability performance of the roof, such as Figure 20 As shown.

[0274] Average angular velocity of top beam offset ;

[0275] angular velocity of top beam offset at any time ;

[0276] Standard deviation of top beam offset angular velocity ;

[0277] In the formula, , This indicates that the top beam offset angle did not reach the expected value. At that time, record Time intervals between adjacent moments, in seconds (s).

[0278] , This indicates that the top beam offset angle did not reach the expected value. At that time, the attitude detection device recorded The angles corresponding to adjacent moments, in degrees (°).

[0279] , … This indicates that the top beam offset angle did not reach the expected value. At any given moment during the descent of the top beam, the angular velocity is expressed in degrees per second. ° / S.

[0280] The angular velocity ω at any moment of the top beam offset is measured. θi Compared with the average Relative deviation (denoted as Δ4), quantifying top plate stability:

[0281]

[0282] Based on the parameter Δ4 value, the top plate condition is divided into three levels, for example:

[0283] This indicates that the angular velocity fluctuation is small, the deformation is uniform, and the top plate has good stability.

[0284] This indicates increased angular velocity dispersion, local disturbances in the top plate, and generally poor stability.

[0285] This indicates that the angular velocity fluctuates drastically, the pressure from the top plate is intense, and the risk of instability is high.

[0286] 4.5 Top Plate Impact Strength G

[0287] (1) Impact strength of the top plate G

[0288] According to the method for evaluating the stability of a mine roof provided by the present invention, when obtaining the stability prediction label, if the roof impact intensity is not greater than the first impact intensity threshold, the roof stability is determined to be level one stable; if the roof impact intensity is greater than the first impact intensity threshold and less than the second impact intensity threshold, the roof stability is determined to be level two stable; if the roof impact intensity is not less than the second impact intensity threshold, the roof stability is determined to be unstable.

[0289] Specifically, the range R between the maximum and minimum impact acceleration values. a and average impact acceleration The ratio of can be used to describe the impact intensity G of the top plate during the impact process.

[0290]

[0291] Based on the impact strength G of the top plate, the standard for classifying the compressive strength of the top plate is as follows:

[0292] when When the pressure from the top plate is relatively low, the risk is small, and an accident is unlikely to occur;

[0293] when At that time, it can be determined that the compressive strength of the roof is average, there is a risk of hazard, and the possibility of an accident is average;

[0294] when At that time, it can be determined that the roof pressure is relatively high, the hazard is relatively serious, and an accident is very likely to occur.

[0295] in and This is a threshold value that can be set according to requirements.

[0296] (2) Average value of safety valve opening time ratio η t

[0297] According to the method for evaluating the stability of a mine roof provided by the present invention, when obtaining stability prediction labels, the average value of the safety valve opening time ratio is obtained; when the average value of the safety valve opening time ratio is not greater than the first average value threshold, the roof stability is determined to be level one stability; when the average value of the safety valve opening time ratio is greater than the first average value threshold and less than the second average value threshold, the roof stability is determined to be level two stability; when the average value of the safety valve opening time ratio is not less than the second average value threshold, the roof stability is determined to be unstable.

[0298] Specifically, the roof impact strength G can be indirectly assessed by monitoring the opening time ratio of safety valves (i.e., the percentage of time the safety valve is open within a specific time period out of the total monitoring time). η tThe higher the value, the more violent the pressure fluctuations on the roof, the more frequent the release of impact energy, and the greater the risk to roof stability.

[0299] According to η t The values ​​are divided into three levels:

[0300] when This indicates that the pressure fluctuations on the top plate are gradual, the impact energy is released and dispersed, and the stability is good.

[0301] when This indicates that localized pressure is concentrated on the roof, periodic impacts are occurring, and there is a risk of strong impacts.

[0302] when This indicates that the energy released during the roof collapse is concentrated, posing an extremely high risk of rock bursts or instantaneous collapse.

[0303] 4.6 Non-uniform top plate pressing P nu

[0304] When the overlying rock fractures and the roof plate impacts the hydraulic support beam, it is necessary to monitor the dynamic changes in the roof pressure. The offset angle θ of the model support beam can reflect the dynamic fluctuations of the roof pressure in real time. Moreover, the offset angle of the beam is more sensitive to pressure changes. It can provide early warning of potential risks through small angle offsets before the support resistance increases significantly, thus reserving time for proactive control.

[0305]

[0306] Where E is the elastic modulus of the top beam material, in GPa; and I is the moment of inertia of the top beam section, in meters. 4 L is the length of the top beam; k is a coefficient determined by the geometric parameters of the support; n is a power parameter, both of which need to be obtained through finite element simulation.

[0307] Pressing P based on the non-uniform top plate nu Criteria for determining roof stability:

[0308] when When the non-uniformity of the roof plate is low, the compressive strength is small, the hazard is small, and accidents are unlikely to occur.

[0309] when At that time, it can be determined that the non-uniformity of the top plate indicates that the compressive strength is generally low, there is a risk of hazard, and the probability of an accident is generally low.

[0310] when When the pressure on the roof is uneven, it can be determined that the pressure intensity is large and the hazard is serious, and an accident is very likely to occur.

[0311] The sum is a critical value that can be set according to requirements.

[0312] 5.1 Comprehensive Analysis and Evaluation of Roof Stability

[0313] For monitoring indicators, a large amount of data from repeated experiments is needed to avoid the randomness of a single experiment and to verify the stability of the roof. However, the repeatability analysis of individual indicators relies entirely on human experience, making efficient and accurate classification and evaluation difficult. Therefore, after completing a small number of individual indicator analyses, a comprehensive roof stability performance evaluation method is provided based on a semi-supervised model algorithm.

[0314] like Figure 21 The diagram shows the algorithm flowchart for evaluating the performance of mine roof slabs according to this invention. This algorithm helps artificial intelligence understand experimental data and adheres to physical principles when processing data, thereby achieving a stability evaluation of the roof slab.

[0315] (1) Original indicators

[0316] The comprehensive stability performance evaluation index of the roof is x(i,j), where i represents the i-th index and j represents the j-th test value, for a total of m.

[0317] Based on the foregoing, the indicators include the periodic pressure step distance D, the dynamic load coefficient N, and the degree of surface breakage of the top plate. The relative deviations of the following parameters are: Δ1 (top plate deformation speed), Δ2 (resistance rise speed at the end of the step), Δ3 (pressure rise speed before safety valve opening), Δ4 (top beam offset angular velocity), G (top plate impact strength), and η (safety valve opening time ratio). t Non-uniform top plate to press P nu There are 10 items in total.

[0318] (2) Data normalization

[0319] S920. Normalize the original indicators to obtain standardized data.

[0320] Specifically, normalization refers to the process of transforming 10 original indicator data of different dimensions and orders of magnitude into a standardized process with unified dimensions and numerical range through mathematical transformation.

[0321]

[0322] In the formula, For the i-th indicator, the j-th test value is the normalized value; For the i-th indicator, the j-th test value; The minimum value among all test values ​​of the i-th indicator; It is the maximum value among all test values ​​of the i-th indicator.

[0323] S930: Input standardized data into the trained neural network model and output stability prediction results.

[0324] The neural network model is trained on a training set, which includes standardized data and stability prediction labels corresponding to multiple monitoring indicators.

[0325] According to the present invention, a method for evaluating the stability of a mine roof slab includes training a neural network model based on a training set. Specifically, this includes: constructing an initial neural network model; determining the consistency constraints of the neural network model based on the convergence time of each monitoring indicator; inputting standardized data into the initial neural network model, training it based on the consistency constraints, and outputting predicted values; obtaining the difference between the predicted values ​​and the back supervision function value; if the difference is less than a preset threshold, temporarily stopping training and saving the trained neural network model; the back supervision function is used to force the neural network to follow physical laws during training and prediction through consistency constraints; performing an objective function test on the predicted values ​​output by the neural network model; iteratively training the neural network model and performing an objective function test based on the neural network model obtained in each iteration; when the number of iterations reaches a preset number, ending the iteration loop and saving the trained neural network model.

[0326] (3) Predicted value y j With index value y i

[0327] The predicted value y is obtained by inputting the indicator test value into the neural network. j With index value y i Predicted value y j Each prediction corresponds one-to-one with the input test values; for example, inputting 100 test values ​​will result in 100 predicted values. The indicator value y... i The neural network predicts the index values ​​in the same step sequence, and selects three of these indices as the completion markers of key actions, which are then used as training material for the neural network to determine the order in which they occur.

[0328] (4) Criteria

[0329] The first criterion is the loop criterion, which repeats the neural network training and objective function testing steps. The loop ends when the number of repetitions reaches k.

[0330] The second criterion is the difference criterion. If the difference between the neural network training function and the back supervision function is less than the criterion, the loop training ends, the model is saved, and the objective function test begins.

[0331]

[0332] Among them, y j * The value of the reverse supervision function, y j These are the predicted values ​​of the neural network training function.

[0333] Reverse supervision function: Select the consistency function of 3 indicators as the semi-supervised function. The purpose of using the consistency function is to force the model to comply with physical rules, such as the safety valve must open after the column retracts.

[0334]

[0335] Where, x 1,j For impact acceleration, x 2,j The relative deviation of the rate of increase of resistance at the end of the step and x 3,j This refers to the ratio of the safety valve opening time.

[0336] (5) Objective function test

[0337] S940. The stability prediction results are scored based on the objective function to obtain the score results, and the roof stability level is determined based on the score results.

[0338] The objective function is designed based on the optimal range distribution law of multiple monitoring indicators.

[0339] Specifically, the objective function is used to calculate y from this model. j The likelihood value is used to test the roof stability, which is the basis for obtaining the roof stability evaluation; essentially, it's a scorer. The higher the likelihood value, the higher the objective function score. For example, a likelihood value of 0.67 (2 / 3) indicates that two out of the three predicted velocity fluctuations meet the target.

[0340] 1) Test method:

[0341] ① Assumption of maximizing the conditional log-likelihood function That is, the standard deviation function is proportional to the distance from the target force to the optimal target force of the support.

[0342]

[0343] ② Solve directly using numerical optimization methods (such as gradient descent).

[0344]

[0345] 2) Objective function:

[0346] In the stability test of the roof slab, the model support has an optimal range. For example, during the test, the initial support force p0, the column rise L, and the final resistance p of the step sequence are set. m Safety valve opening pressure p k Standard parameters such as the reduction in volume Sn under the column fall within this range. Within this optimal range, most target values ​​deviate relatively close, while a few deviate significantly. If the test is not within the optimal range, a few target values ​​deviate relatively close, while most deviate significantly.

[0347] Based on this standard, design an objective function that conforms to this rule, such as... Figure 20 As shown.

[0348] The objective function should satisfy the following conditions: when the target force is within the optimal range of stent performance, the target value of the objective function is relatively concentrated; when the target force is outside the optimal range of stent performance, the target force of the objective function is relatively dispersed.

[0349] The objective function is obtained by fitting a model, which follows a normal distribution with a constant mean and varying variance.

[0350]

[0351] Where μ is the optimal target force of the stent, (p-μ) is the distance from the target force to the optimal target force of the stent, and σ(p-μ) is the standard deviation, which is positively correlated with the distance from the target force to the optimal target force of the stent.

[0352] 5.2 Comprehensive Grade Determination

[0353] According to the method for evaluating the stability of a mine roof provided by the present invention, the roof stability level is determined based on the scoring results, specifically including: the stability level includes level one, level two, and level three; if the objective function score is not less than the first threshold and all consistency constraints of the neural network model are satisfied, the roof stability level is determined to be level one; if the objective function score is not less than the second threshold and all consistency constraints of the neural network model are satisfied, the roof stability level is determined to be level two; if the objective function score is less than the third threshold or any consistency constraint of the neural network model is violated, the roof stability level is determined to be level three.

[0354] Table 1

[0355]

[0356] There is a key constraint: if the likelihood value is less than 0.33, the roof is directly determined to be unstable.

[0357] 5.3 Model Adjustment Methods

[0358] The functions used in the algorithm can all be dynamically adjusted according to requirements, for example:

[0359] 1) Select another function that is positively correlated with the distance from the target force to the optimal target force on the support as the variance function;

[0360] 2) Select other x i y was calculated j The model;

[0361] 3) Choose another backpropagation function to represent x corresponding to this index.

[0362] The following describes the mine roof stability evaluation device provided by the present invention. The mine roof stability evaluation device described below can be referred to in correspondence with the mine roof stability evaluation method described above.

[0363] like Figure 23 The image shows a mine roof stability evaluation device provided by the present invention, comprising:

[0364] The indicator acquisition module 2310 is used to acquire multiple monitoring indicators of the pressure on the top of the mine roof and determine the original indicators based on the monitoring indicators. The monitoring indicators include the end resistance of the step sequence, the average resistance of the step sequence, the opening pressure of the safety valve, the amount of roof deformation, the impact acceleration, and the offset angle of the roof beam.

[0365] Standardization module 2320 is used to normalize the original indicators to obtain standardized data;

[0366] The model prediction module 2330 is used to input standardized data into a trained neural network model and output stability prediction results. The neural network model is trained based on a training set, which includes standardized data and stability prediction labels corresponding to multiple monitoring indicators.

[0367] The performance evaluation module 2340 is used to score the stability prediction results based on the objective function to obtain the score results, and to determine the roof stability level based on the score results; the objective function is designed based on the optimal range distribution law of multiple monitoring indicators.

[0368] Figure 24 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 24As shown, the electronic device may include: a processor 2410, a communication interface 2420, a memory 2430, and a communication bus 2440. The processor 2410, communication interface 2420, and memory 2430 communicate with each other via the communication bus 2440. The processor 2410 can call logical instructions in the memory 2430 to execute a mine roof stability evaluation method. This method includes: acquiring multiple monitoring indicators of mine roof pressure; determining original indicators based on the monitoring indicators; the monitoring indicators include step-end resistance, step-average resistance, safety valve opening pressure, roof deformation, impact acceleration, and roof beam offset angle; normalizing the original indicators to obtain standardized data; inputting the standardized data into a trained neural network model and outputting stability prediction results; the neural network model is trained based on a training set, which includes standardized data corresponding to multiple monitoring indicators and stability prediction labels; scoring the stability prediction results based on an objective function to obtain a score result; and determining the roof stability level based on the score result; the objective function is designed based on the optimal range distribution law of multiple monitoring indicators.

[0369] Furthermore, the logical instructions in the aforementioned memory 2430 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0370] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the mine roof stability evaluation method provided by the above methods. The method includes: acquiring multiple monitoring indicators of mine roof pressure; determining original indicators based on the monitoring indicators; the monitoring indicators include step-end resistance, step-average resistance, safety valve opening pressure, roof deformation, impact acceleration, and roof beam offset angle; normalizing the original indicators to obtain standardized data; inputting the standardized data into a trained neural network model and outputting stability prediction results; the neural network model is trained based on a training set, which includes standardized data corresponding to multiple monitoring indicators and stability prediction labels; scoring the stability prediction results based on an objective function to obtain a scoring result; and determining the roof stability level based on the scoring result; the objective function is designed based on the optimal range distribution law of multiple monitoring indicators.

[0371] In another aspect, the present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon. When executed by a processor, the computer program implements the method for evaluating the stability of the mine roof provided by the above methods. The method includes: acquiring multiple monitoring indicators of the mine roof pressure; determining original indicators based on the monitoring indicators; the monitoring indicators include step-end resistance, step-average resistance, safety valve opening pressure, roof deformation, impact acceleration, and roof beam offset angle; normalizing the original indicators to obtain standardized data; inputting the standardized data into a trained neural network model and outputting stability prediction results; the neural network model is trained based on a training set, which includes standardized data corresponding to multiple monitoring indicators and stability prediction labels; scoring the stability prediction results based on an objective function to obtain a score result; and determining the roof stability level based on the score result; the objective function is designed based on the optimal range distribution law of multiple monitoring indicators.

[0372] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0373] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of various embodiments or some parts of embodiments.

[0374] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for evaluating the stability of a mine roof, characterized in that, include: Multiple monitoring indicators of mine roof pressure are obtained, and raw indicators are determined based on the monitoring indicators; the monitoring indicators include end resistance of step sequence, average resistance of step sequence, safety valve opening pressure, roof deformation, impact acceleration, and roof beam offset angle. The original indicators are normalized to obtain standardized data; The standardized data is input into a trained neural network model, which outputs stability prediction results. The neural network model is trained based on a training set, which includes standardized data and stability prediction labels corresponding to multiple original indicators. The original indicators include periodic pressure step distance, dynamic load coefficient, degree of roof surface breakage, roof deformation rate, resistance rise rate at the end of the step, pressure rise rate before safety valve opening, roof beam offset angular velocity, roof impact strength, safety valve opening time ratio, and non-uniform roof pressure. The stability prediction results are scored based on the objective function to obtain a score result, and the roof stability level is determined based on the score result. The objective function is designed based on the optimal range distribution law of multiple monitoring indicators; The neural network model is trained based on the training set, specifically including: constructing an initial neural network model; the consistency constraints of the neural network model are determined based on the convergence time of each monitoring indicator; The standardized data is input into the initial neural network model, and training is performed based on the consistency constraint to output predicted values. The difference between the predicted values ​​and the back supervision function values ​​is obtained. If the difference is less than a preset threshold, training is temporarily stopped and the trained neural network model is saved. The back supervision function is used to force the neural network to follow physical laws during training and prediction through the consistency constraint. The predicted values ​​output by the neural network model are tested for the objective function. The neural network model is trained iteratively, and the objective function is tested based on the neural network model obtained in each iteration. When the number of iterations reaches a preset number, the iteration loop ends and the trained neural network model is saved.

2. The method for evaluating the stability of a mine roof slab according to claim 1, characterized in that, The monitoring indicators of the top plate are obtained directly through sensors.

3. The method for evaluating the stability of a mine roof slab according to claim 2, characterized in that, The process of determining the original indicators based on the monitoring indicators specifically includes: The step distance is reduced by determining the cycle based on the resistance at the end of the step sequence or the average resistance of the step sequence. The dynamic load factor is determined based on the step average resistance during the period of cyclic pressure on the roof and the period of non-pressure on the roof. The degree of surface breakage of the roof plate is determined based on the amount of deformation of the roof plate. Determine the deformation rate of the top plate based on the amount of deformation. Determine the rate of increase of the resistance at the end of the step sequence based on the resistance at the end of the step sequence; Determine the rate of pressure rise before the safety valve opens based on the safety valve's opening pressure. Determine the angular velocity of the top beam offset based on the top beam offset angle; The impact strength of the top plate is determined based on the impact acceleration. The average ratio of safety valve opening time is determined based on the safety valve opening pressure. The non-uniform top plate is determined based on the offset angle of the top beam.

4. The method for evaluating the stability of a mine roof slab according to claim 3, characterized in that, The method of determining the period based on the end-of-step resistance or the average step resistance to reduce the step distance specifically includes: Obtain the average and range of the step-by-step average resistance; The cycle is determined based on the average value and range of the average resistance in the step sequence, which is the main indicator for judging pressure. When the average resistance of the step sequence is greater than the main discrimination index of periodic pressure, it is determined that periodic pressure has occurred on the top plate, and the periodic pressure step distance is obtained. or Obtain the average and range of the resistance at the end of the step sequence; The cycle is determined by the average value and range of the resistance at the end of the step sequence to assist in the judgment of indicators. When the resistance at the end of the step is greater than the periodic pressure auxiliary judgment index, it is determined that the top plate has experienced periodic pressure, and the periodic pressure step distance is obtained.

5. The method for evaluating the stability of a mine roof slab according to claim 3, characterized in that, The determination of the dynamic load factor based on the step-by-step average resistance during the period of roof cyclic pressing and the period of roof non-pressing specifically includes: The ratio of the average resistance during the period of periodic pressure on the roof to that during the period of non-pressure on the roof is obtained as the dynamic load factor.

6. The method for evaluating the stability of a mine roof slab according to claim 3, characterized in that, The determination of the degree of surface breakage of the top plate based on the deformation of the top plate specifically includes: Obtain the range and average value of the top plate deformation; The ratio of the range to the average value of the deformation of the top plate is determined as the degree of surface breakage of the top plate.

7. The method for evaluating the stability of a mine roof slab according to claim 2, characterized in that, When obtaining the stability prediction label, the method further includes: When the periodic pressure step distance is not less than the first pressure step distance threshold, the top plate stability is determined to be at level one stability. When the periodic pressure step distance is less than the first pressure step distance threshold and greater than the second pressure step distance threshold, the top plate stability is determined to be level two stability. When the periodic pressure step distance is not greater than the second pressure step distance threshold, the stability of the top plate is determined to be unstable.

8. The method for evaluating the stability of a mine roof slab according to claim 2, characterized in that, When obtaining the stability prediction label, the method further includes: When the dynamic load factor is not greater than the first dynamic load factor threshold, the stability of the top plate is determined to be Level 1 stability; When the dynamic load factor is greater than the first dynamic load factor threshold and less than the second dynamic load factor threshold, the stability of the top plate is determined to be at level two stability. When the dynamic load factor is not less than the second dynamic load factor threshold, the stability of the top plate is determined to be unstable.

9. The method for evaluating the stability of a mine roof slab according to claim 2, characterized in that, When obtaining the stability prediction label, the method further includes: When the degree of breakage of the top plate surface is not greater than the first surface breakage threshold, the stability of the top plate is determined to be level one stability; When the degree of surface breakage of the top plate is greater than the first surface breakage threshold and less than the second surface breakage threshold, the stability of the top plate is determined to be level two stability. When the degree of breakage of the top plate surface is not less than the second surface breakage threshold, the stability of the top plate is determined to be unstable.

10. The method for evaluating the stability of a mine roof slab according to claim 2, characterized in that, When obtaining the stability prediction label, the method further includes: The first relative error between the adjacent time displacement difference method and the displacement fitting curve differentiation method is determined based on the deformation rate of the top plate. The first relative deviation between the maximum value and the average value of the top plate deformation rate is determined based on the value of the first relative error; When the first relative deviation is not greater than the first relative deviation threshold, the stability of the top plate is determined to be at level one stability. When the first relative deviation is greater than the first relative deviation threshold and less than the second relative deviation threshold, the stability of the top plate is determined to be level two stability. When the first relative deviation is not less than the second relative deviation threshold, the stability of the top plate is determined to be unstable.

11. The method for evaluating the stability of a mine roof slab according to claim 2, characterized in that, When obtaining the stability prediction label, the method further includes: The second relative error between the adjacent time displacement difference method and the displacement fitting curve differentiation method is determined based on the rising speed of the resistance at the end of the step sequence. The second relative deviation between the maximum value and the average value of the resistance rise rate at the end of the step sequence is determined based on the value of the second relative error. When the second relative deviation is not greater than the third relative deviation threshold, the stability of the top plate is determined to be at level one stability. When the second relative deviation is greater than the third relative deviation threshold and less than the fourth relative deviation threshold, the top plate stability is determined to be level two stability. When the second relative deviation is not less than the fourth relative deviation threshold, the stability of the top plate is determined to be unstable.

12. The method for evaluating the stability of a mine roof slab according to claim 2, characterized in that, When obtaining the stability prediction label, the method further includes: The third relative error between the adjacent time displacement difference method and the displacement fitting curve differentiation method is determined based on the pressure rise rate before the safety valve opens. The third relative error is used to determine the third relative deviation between the maximum value and the average value of the pressure rise rate before the safety valve opens; When the third relative deviation is not greater than the fifth relative deviation threshold, the stability of the top plate is determined to be at level one stability. When the third relative deviation is greater than the fifth relative deviation threshold and less than the sixth relative deviation threshold, the top plate stability is determined to be level two stability. When the third relative deviation is not less than the sixth relative deviation threshold, the stability of the top plate is determined to be unstable.

13. The method for evaluating the stability of a mine roof slab according to claim 2, characterized in that, When obtaining the stability prediction label, the method further includes: When the impact strength of the top plate is not greater than the first impact strength threshold, the stability of the top plate is determined to be level one stability; When the impact intensity of the top plate is greater than the first impact intensity threshold and less than the second impact intensity threshold, the stability of the top plate is determined to be level two stability. When the impact intensity of the top plate is not less than the second impact intensity threshold, the stability of the top plate is determined to be unstable.

14. The method for evaluating the stability of a mine roof slab according to claim 2, characterized in that, When obtaining the stability prediction label, the method further includes: Obtain the average ratio of safety valve opening time; When the average opening time ratio of the safety valve is not greater than the threshold value of the first time ratio, the stability of the top plate is determined to be at level one stability. When the average value of the safety valve opening time ratio is greater than the first average time ratio threshold and less than the second average time ratio threshold, the top plate stability is determined to be at level two stability. When the average value of the safety valve opening time ratio is not less than the threshold value of the second average time ratio, the stability of the top plate is determined to be unstable.

15. The method for evaluating the stability of a mine roof slab according to claim 1, characterized in that, The process of determining the top plate stability level based on the scoring results specifically includes: The stability levels include Level 1, Level 2, and Level 3; If the objective function score is not less than the first threshold and all consistency constraints of the neural network model are met, then the stability level of the roof is determined to be Level 1. If the objective function score is not less than the second threshold and all consistency constraints of the neural network model are met, then the stability level of the roof is determined to be level two. If the objective function score is less than the third threshold or violates any consistency constraint of the neural network model, the stability level of the roof is determined to be level three.

16. A device for evaluating the stability of a mine roof, characterized in that, include: The indicator acquisition module is used to acquire multiple monitoring indicators of the pressure on the top of the mine roof and determine the original indicators based on the monitoring indicators. The monitoring indicators include the end resistance of the step sequence, the average resistance of the step sequence, the opening pressure of the safety valve, the amount of roof deformation, the impact acceleration, and the offset angle of the roof beam. The standardization module is used to normalize the original indicators to obtain standardized data; The model prediction module is used to input the standardized data into a trained neural network model and output stability prediction results. The neural network model is trained based on a training set, which includes standardized data and stability prediction labels corresponding to multiple original indicators. The original indicators include periodic pressure step distance, dynamic load coefficient, degree of roof surface breakage, roof deformation rate, resistance rise rate at the end of the step, pressure rise rate before safety valve opening, roof beam offset angular velocity, roof impact strength, safety valve opening time ratio, and non-uniform roof pressure. The performance evaluation module is used to score the stability prediction results based on the objective function to obtain a score result, and to determine the stability level of the top plate based on the score result; The objective function is designed based on the optimal range distribution law of multiple monitoring indicators; The device is also used to: train the neural network model according to the training set, specifically including: constructing an initial neural network model; the consistency constraints of the neural network model are determined based on the convergence time of each monitoring indicator; The standardized data is input into the initial neural network model, and training is performed based on the consistency constraint to output predicted values. The difference between the predicted values ​​and the back supervision function values ​​is obtained. If the difference is less than a preset threshold, training is temporarily stopped and the trained neural network model is saved. The back supervision function is used to force the neural network to follow physical laws during training and prediction through the consistency constraint. The predicted values ​​output by the neural network model are tested for the objective function. The neural network model is trained iteratively, and the objective function is tested based on the neural network model obtained in each iteration. When the number of iterations reaches a preset number, the iteration loop ends and the trained neural network model is saved.

17. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the mine roof stability evaluation method as described in any one of claims 1 to 15.

18. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the mine roof stability evaluation method as described in any one of claims 1 to 15.

19. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the mine roof stability evaluation method as described in any one of claims 1 to 15.

Images