Gas burst parameter prediction and borehole design method based on back propagation neural network

Abstract

The present application provides a gas blasting parameter prediction and hole arrangement design method based on a back propagation neural network, comprising the following steps: S1, obtaining physical and mechanical parameters of rock mass in a blasting area; S2, carrying out single-hole gas blasting test to determine an optimal burden; S3, repeating steps S1 and S2 at different sites, counting corresponding data, and establishing a database; S4, constructing and training a back propagation neural network, and predicting reasonable gas blasting parameters by inputting training data; and S5, optimizing hole arrangement design according to the network prediction result, and determining hole depth, hole distance and hole spacing. The present application can accurately predict blasting parameters, reasonably design hole arrangement parameters in combination with rock mass parameters and site conditions, significantly reduce rock breaking cost, improve rock breaking efficiency, and further optimize economic benefits of blasting operation.

Description

Technical Field

[0001] This invention belongs to the field of rock mass engineering blasting construction, specifically involving a method for predicting gas blasting parameters and designing borehole layout based on a backpropagation neural network. Background Technology

[0002] Explosive blasting has advantages such as high efficiency and economy, and is widely used in rock excavation for water conservancy, hydropower, transportation, energy, and municipal engineering projects. However, the disadvantages of explosive blasting are also very prominent. The use of explosives carries high safety risks and requires strict management and control. The resulting blast shock waves can cause significant vibration, noise, flying rocks, and damage to the rock mass in the reserved area, leading to a series of environmental problems and engineering disasters.

[0003] To control vibration, flyrock, and damage to surrounding rock, and to overcome the drawbacks of explosive blasting, gas blasting technologies such as CO2 phase change blasting are often used as alternatives to explosives in rock-breaking excavation operations. CO2 blasting involves rapidly heating carbon dioxide within a fracturing tube, causing its pressure to rise rapidly, and then suddenly releasing it, creating an impact force on the surrounding rock mass, thus achieving rock-breaking. Because the activator materials used in early CO2 phase change blasting technologies were hazardous chemicals, certain safety risks remained. Some manufacturers improved the activator formulas, thereby avoiding the use of hazardous chemicals. Furthermore, driven by market demand, supercritical fluid impact rock-breaking technology emerged, as described in Chinese invention patent CN118463736A, "Method for Open-cut Excavation of Rock Based on Supercritical Fluid Impact Rock-Breaking Equipment." Supercritical fluid impact rock-breaking technology uses the rapid release of high-pressure gas to impact and break rock, similar in principle to CO2 phase change blasting. In this invention, technologies with similar principles, such as CO2 blasting and supercritical fluid impact rock-breaking technology, are collectively referred to as gas blasting.

[0004] To date, the design of borehole network parameters in actual gas blasting operations is still largely based on experience, exhibiting a degree of blindness. Chinese invention patent CN112364489B, "A Carbon Dioxide Blasting Construction Method for Controlling Bedrock Damage and Vibration Effects," adjusts construction parameters according to damage and vibration control requirements, rather than designing borehole parameters with the goal of improving rock-breaking efficiency and reducing rock-breaking costs.

[0005] Low profit margins are a prominent feature of the construction industry. In gas blasting construction, if reasonable hole layout parameters can be determined based on rock mass parameters and site conditions, rock breaking costs will be significantly reduced, greatly improving the market competitiveness of gas blasting technology. Summary of the Invention

[0006] The purpose of this invention is to overcome the shortcomings of the prior art and provide a method for predicting gas explosion parameters and designing apertures based on backpropagation neural networks, so as to solve the problems mentioned in the above-mentioned technical background.

[0007] To achieve the above objectives, the present invention provides the following technical solution:

[0008] A method for predicting gas explosion parameters and designing aperture layout based on backpropagation neural networks includes the following steps:

[0009] Step S1: Obtain the physical and mechanical parameters of the rock mass in the blasting area, including: the point load strength I of the rock block. s(50) Wave impedance ρV of the rock block pr Rock mass integrity index K v The influence coefficient of the attitude of the main structural planes, K5;

[0010] Rock mass is a geological body composed of structural planes and the enclosed rock blocks, possessing complex physical and mechanical properties. The selected physical and mechanical parameters should significantly influence the gas blasting effect while also being easy to measure. The rock mass parameters selected in this invention include: the point load strength I of the rock blocks. s(50) Wave impedance ρV of the rock block pr Rock mass integrity index K v The influence coefficient of the attitude of the main structural planes, K5. Where ρ is the density of the rock block, and K is the rock mass integrity index. v Calculate according to formula (1):

[0011] K v =(V pm / V pr ) 2 (8)

[0012] In the formula, V pm It is the elastic longitudinal wave velocity of the rock mass, V pr It is the elastic longitudinal wave velocity of the rock block. The influence coefficient K5 of the attitude of the main structural plane is calculated according to formula (2):

[0013] K5 = F1 × F2 × F3 (9)

[0014] In the formula, F1 represents the influence of the angle between the dip of the structural surface and the dip of the broken surface, F2 represents the influence of the dip angle of the structural surface, and F3 represents the influence of the difference between the dip angle of the structural surface and the dip angle of the slope surface. F1, F2, and F3 are taken according to Table 5.3.2-3 in the "Engineering Rock Mass Classification Standard" (GB / T50218-2014).

[0015] Step S2: Conduct multiple single-hole gas blasting tests in the same site, measure and record the rock breaking volume of each test, and determine the optimal resistance line and other key blasting parameters under the corresponding site conditions, including blasting energy, unit consumption and gas blasting energy per hole length.

[0016] Further, step S2 includes the following steps:

[0017] Step S201: Conduct multiple single-hole gas blasting rock-breaking tests at the same site, keeping the test conditions consistent each time, adjusting only the resistance line W. Determine the maximum rock-breaking volume V by statistically analyzing the rock-breaking volume V of each test. O The corresponding optimal resistance line W under these site conditions O

[0018] Step S202: Calculate the explosion energy E of the single-hole test according to the explosion energy calculation formula for compressed gas and water vapor containers. g :

[0019]

[0020] In the formula, k is the adiabatic index of carbon dioxide, taken as 1.295; p is the burst pressure; p atm It is atmospheric pressure; V0 is the initial gas volume of a single explosion, which is the volume of the fracturing device multiplied by the number of single-hole fracturing devices.

[0021] Step S203: Based on the blast energy E g Calculate the optimal unit consumption e under the corresponding rock mass properties and site conditions. O :

[0022]

[0023] And the gas explosion energy λ per meter along the borehole:

[0024]

[0025] In the formula, H is the hole depth and T is the plugging length.

[0026] Step S3: Repeat steps S1 and S2 at blasting sites under different geological conditions, accumulate data through multiple experiments, and statistically analyze the relevant data and key blasting parameters obtained from the experiments to establish a database of gas blasting rock breaking operations under different geological conditions; the database list is shown in Table 1, where α represents the slope inclination angle of the bench.

[0027] Table 1. Gas Explosion Database List

[0028] Site No. I s(50) ]] pV pr ]]> K v ]]> [K5] α λ [WC O ]]> e O ]]> 1 2 ... N

[0029] Step S4: Construct and train a gas explosion parameter prediction model based on a backpropagation neural network. Input training data includes physical and mechanical parameters and experimental data. Output predicted gas explosion parameters. Optimize the neural network structure through the backpropagation algorithm to improve prediction accuracy.

[0030] Step S401: Based on the characteristics and correlations of the gas explosion parameters, determine the topology of the backpropagation neural network and construct a preliminary neural network model;

[0031] A backpropagation neural network is established, including an input layer, an output layer, and a hidden layer. Preferably, the backpropagation neural network model established in this invention includes one hidden layer. The rock block point load intensity I... s(50) Rock block impedance ρV pr Rock mass integrity index K v The six parameters—K5 (the influence coefficient of the main structural plane attitude), α (the slope angle of the step), and λ (the energy of gas explosion per meter)—are used as input indicators to determine the reasonable resistance line W. O and reasonable unit consumption e O For the output metrics, the input layer has 6 neurons and the output layer has 2 neurons. The hidden layer neurons extract and store the inherent patterns from the samples. Each hidden layer neuron has several weights, and each weight is a parameter that enhances the network's mapping ability. If the number of hidden layer neurons is too small, the network's ability to extract information from the samples will be poor, insufficient to summarize and represent the patterns in the training samples; if the number of hidden layer neurons is too large, it may learn and memorize irregular content from the samples, leading to an "overfitting" problem, which actually reduces the network's generalization ability. Therefore, the number of hidden layer neurons depends on the number of training samples, the level of sample noise, and the complexity of the patterns contained in the samples. According to Kolmogolov's theorem, the number of hidden layer units in this invention is 2m+1, where m is the number of training samples.

[0032] Step S402: Select an activation function suitable for predicting gas explosion parameters and set initial weights;

[0033] In backpropagation neural network models, the activation function f is usually a differentiable monotonically increasing function. In this invention, the activation function of the backpropagation neural network model is a sigmoid-type logarithmic function logsig. The advantage of choosing the sigmoid function is that any input data can be transformed into a number between (0, +1). The activation function of the last layer of neurons is a purelin-type function.

[0034] For backpropagation neural networks, different initial weights result in different training outcomes each time, due to the large number of local minima on the error surface. In this invention, random numbers between (-1, 1) are used as the initial values.

[0035] Step S403: Perform normalization preprocessing on the sample data, including normalization of the input samples and outlier removal;

[0036] To standardize the data and make the predictions more reasonable and reliable, before using actual engineering data as samples to train the neural network, the significant differences in the numerical values ​​between the data points can increase the difficulty of network training and reduce accuracy. To eliminate the influence of different dimensions among variables, normalization transformation should be performed. Sample data can only be used by the model and to maximize its performance after preprocessing. In this invention, the original data is normalized according to the following method:

[0037]

[0038] In the formula, X represents the original data; X max X min T represents the maximum and minimum values ​​of the original data; T represents the target data. max T min The maximum and minimum values ​​of the target data are set to 0.8 and 0.2 respectively.

[0039] Step S404: Using the normalized sample data, train the neural network using the backpropagation algorithm, and optimize the weights through multiple iterations to form a neural network model for predicting gas explosion parameters;

[0040] The neural network is trained using a feedforward multilayer neural network error backpropagation learning algorithm, with the training function being a Bayesian regularized training function `trainbr`. The parameters that need to be set in this algorithm are the network's expected error, learning rate, and number of training iterations. When the error is less than or equal to a set value, the network weights and threshold are saved, and training ends.

[0041] The trained neural network model can then be used to predict the optimal resistance line W in the blasting area. O and reasonable optimal unit consumption e O .

[0042] Step S5: Based on the neural network model trained in step S4, use the model to optimize the hole layout parameters for gas blasting, including the reasonable setting of resistance line, hole spacing, row spacing, and hole depth.

[0043] Furthermore, step S5 specifically includes the following steps:

[0044] Step S501: Obtain rock mass parameters in the blasting area, including the point load strength I of rock blocks on the step slope. s(50) Rock block impedance ρV pr Rock mass integrity index K v The main structural plane attitude influence coefficient K5, the step slope dip angle α, and the step height H B ;

[0045] Step S502: Determine the hole depth H and the appropriate fracturing device model based on the step height, and calculate the gas explosion energy λ per meter according to formula (5); In gas explosion, the fracturing device has a fixed model, and the length and explosion energy of each model are fixed. The hole depth H is generally an integer multiple of the fracturing device length plus the plugging length, and the hole depth H is less than or close to the step height H. B Once the type and depth of the fracturing device are determined, the gas explosion energy λ per meter can be determined according to equation (5).

[0046] Step S503: Input the parameters into the trained neural network model to predict the reasonable optimal resistance line W. O and reasonable optimal unit consumption e O The point load strength of the rock block I s(50) Rock block impedance ρV pr Rock mass integrity index K v The six parameters—K5 (the influence coefficient of the main structural plane attitude), α (the slope angle of the step), and λ (the energy of the gas explosion per meter)—are input into a trained backpropagation neural network to obtain the reasonable resistance line W. O and reasonable unit consumption e O ;

[0047] Step S504: Determine the hole layout parameters and adopt a triangular hole layout, wherein the distance from the first row of holes to the edge of the step is W. O The spacing between holes in the same row is:

[0048]

[0049] In the formula, b is the row spacing. Because gas blasting requires a high degree of freedom in rock breaking, it is generally carried out in actual blasting by detonating row by row or hole by hole. After the first row of holes is detonated, a free face is created for the second row. Therefore, the row spacing is also defined as b = W. O Meanwhile, when conditions permit, triangular hole layout should be used as much as possible, that is, the holes in adjacent rows should be staggered.

[0050] Compared with the prior art, the present invention has the following beneficial effects:

[0051] This invention employs a gas blasting parameter prediction and borehole design method based on a backpropagation neural network. This method accurately predicts blasting parameters and automatically optimizes borehole design, improving blasting effectiveness, resource utilization efficiency, and reducing human error. The method is highly adaptable and can effectively enhance engineering safety and economic benefits. Attached Figure Description

[0052] Figure 1 A diagram of the backpropagation neural network structure constructed according to an embodiment of the present invention;

[0053] Figure 2This example shows the neural network structure and execution process displayed during program runtime.

[0054] Figure 3 The example shows the trend of the mean square error of the neural network as a function of the number of training iterations. Detailed Implementation

[0055] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0056] The purpose of this invention is to overcome the shortcomings of existing technologies and provide a reasonable parameter prediction model and borehole parameter design method for gas blasting based on backpropagation neural networks. This method determines reasonable borehole parameters according to rock mass parameters and site conditions, reduces rock breaking costs, and improves the market competitiveness of gas blasting technology.

[0057] This invention provides a reasonable parameter prediction model for gas explosion based on a backpropagation neural network, comprising the following steps:

[0058] Step S1: Obtain the physical and mechanical parameters of the rock mass in the blasting area, including: the point load strength I of the rock block. s(50) Wave impedance ρV of the rock block pr Rock mass integrity index K v The influence coefficient of the attitude of the main structural planes, K5;

[0059] The average point load strength I of a certain calcareous breccia site. s(50) The average impedance of the rock block is ρV, which is 2.828 MPa. pr It is 11.693×10 6 Pa·s / m, rock mass integrity index K v The value is 0.38; the difference between the dip angle of the control structure surface and the dip angle of the step slope within the site is greater than 10°. According to Table 5.3.2-3 of the "Engineering Rock Mass Classification Standard" (GB / T 50218-2014), F3 is 0, so K5=F1×F2×F3=0.

[0060] Step S2: Conduct multiple single-hole gas blasting tests in the same site, measure and record the rock breaking volume of each test, and determine the optimal resistance line and other key blasting parameters under the corresponding site conditions, including blasting energy, unit consumption and gas blasting energy per hole length.

[0061] The platform height is approximately 10m, with an inclination angle of 85°. A Φ89×4×1200 type disposable fracturing device was used, with a single-unit gas explosion energy of 658.4kJ. Single-hole blasting tests were conducted using boreholes 2.5m deep, with one fracturing device placed in each hole and a plugging length of 1.3m. Therefore, the gas explosion energy per meter is λ = 548.67kJ / m. Five single-hole gas explosion tests were performed, with resistance lines of 0.9m, 1.1m, 1.3m, 1.5m, and 1.7m. The largest rock breaking volume (6.5m) was observed at a resistance line of 1.5m. 3 Therefore, the reasonable resistance line for this region is W. O =1.5m, the reasonable unit consumption is e O =101.3kJ / m 3 .

[0062] Step S3: Repeat steps S1 and S2 at blasting sites under different geological conditions, accumulate data through multiple experiments, and statistically analyze the relevant data and key blasting parameters obtained from the experiments to establish a database of gas blasting rock breaking operations that includes different geological conditions.

[0063] Following the requirements described in steps S1 and S2, gas blasting rock breaking operations were carried out under different rock mass and site conditions. The corresponding data were collected, and a sample database was established as shown in Table 2.

[0064] Table 2. Gas Explosion Sample Database

[0065] No. I s(50) (MPa) ρV pr (MPa·s / m)]]> K v ]]> [K5] α(°) λ (kJ / m) W O (m)]]> e O (kJ / m 3 )]]> 1 2.83 11.7 0.38 0 85 548.7 1.5 101.3 2 5.92 13.5 0.22 0.8 87 902.7 2.0 150.5 3 7.63 15.6 0.65 0.32 80 366.1 1.5 252.6 4 6.14 16.3 0.74 0 80 902.7 1.1 548.3 5 7.69 13.6 0.55 0.17 80 366.1 1.4 236.5 6 8.32 12.9 0.43 0.17 80 366.1 1.3 284.9 7 10.2 13.5 0.54 0 80 902.7 1.1 532.1 8 11.6 15.6 0.46 0.68 85 902.7 1.6 317.1 9 10.3 15.0 0.48 0.14 90 902.7 1.3 430.1 10 8.95 14.9 0.58 0 78 548.7 1.0 419.3 11 8.36 14.8 0.66 0 82 548.7 1.1 413.9 12 7.55 15.0 0.65 0.08 82 366.1 1.2 295.6 13 9.78 15.8 0.28 0.32 84 366.1 1.5 198.7 14 9.99 11.4 0.48 0 81 902.7 1.1 484.1 15 6.27 12.3 0.37 0.08 78 548.7 1.2 303.4 16 9.53 13.8 0.69 0.3 88 902.7 1.4 400.9

[0066] Step S4: Construct and train a gas explosion parameter prediction model based on a backpropagation neural network. The input training data includes physical and mechanical parameters and experimental data. The output is the predicted gas explosion parameters. The neural network structure is optimized through the backpropagation algorithm to improve the prediction accuracy. Specifically, this includes the following steps:

[0067] Step S401: Based on the characteristics and correlations of the gas explosion parameters, determine the topology of the backpropagation neural network and construct a preliminary neural network model;

[0068] The first 12 samples in Table 2 are used as training samples, and the last 4 samples are used as test samples to verify the prediction accuracy of the neural network. Figure 1 As shown, a backpropagation neural network is established, consisting of one input layer, one hidden layer, and one output layer. The input layer contains six nodes, each corresponding to a point load intensity I of the rock block. s(50) Rock block impedance ρV pr Rock mass integrity index K v The six parameters are: the main structural plane attitude influence coefficient K5, the step slope dip angle α, and the gas explosion energy per meter λ; the output layer contains two nodes, corresponding to the reasonable resistance line W.O and reasonable unit consumption e O The number of training samples is 12, so the number of hidden layer nodes is set to 12×2+1=25.

[0069] Step S402: Select an activation function suitable for predicting gas explosion parameters and set initial weights;

[0070] The activation function for the intermediate layer neurons in the backpropagation neural network model is a sigmoid-type logarithmic function (logsig), while the activation function for the last layer neurons is a purelin-type function. The initial weights of the network are random numbers between (-1, 1).

[0071] Step S403: Perform normalization preprocessing on the sample data, including normalization of the input samples and outlier removal;

[0072] To eliminate the influence of dimensions between variables, the variables should be normalized. Sample data must be preprocessed before it can be used by the model and to maximize its performance. The original data is normalized according to equation (6).

[0073] Step S404: Using the normalized sample data, train the neural network using the backpropagation algorithm, and optimize the weights through multiple iterations to form a neural network model for predicting gas explosion parameters;

[0074] The neural network is trained using a feedforward multilayer neural network error backpropagation learning algorithm, with the Bayesian regularized training function `trainbr`. The parameters that need to be set in this algorithm are the network's expected error, learning rate, and number of training iterations; in this embodiment, these are set to 1.0 × 10⁻⁶. -10 0.01 and 10 11 When the error is less than or equal to the set value, the network weights and threshold are saved, and training ends. The network structure and operation process displayed during program execution are as follows: Figure 2 As shown, the mean squared error gradually converges with the increase of training iterations during the training process, as... Figure 3 As shown.

[0075] The prediction accuracy of the neural network model was not tested. The measured results and prediction results of the last four samples in Table 2 are listed in Table 3. It can be seen from Table 3 that the prediction accuracy of the model is relatively high.

[0076] Table 3 Comparison of Predicted Results and Measured Results

[0077]

[0078]

[0079] The trained neural network can then be used to predict the optimal resistance line W in the blasting area. O and reasonable optimal unit consumption e O .

[0080] Step S5: Based on the neural network model trained in step S4, use the model to optimize the hole layout parameters for gas blasting, including the reasonable setting of resistance line, hole spacing, row spacing, and hole depth.

[0081] Step S501: Obtain rock mass parameters in the blasting area, including the point load strength I of rock blocks on the step slope. s(50) Rock block impedance ρV pr Rock mass integrity index K v The main structural plane attitude influence coefficient K5, the step slope dip angle α, and the step height H B ;

[0082] Taking a gas blasting site in a mining area as an example, the following results were obtained through experiments and measurements: I0, the point load strength of the rock blocks in the blasting area. s(50) 6.0 MPa, rock block wave impedance ρV pr The rock mass integrity index K is 12.8 MPa·s / m. v The coefficient of influence of the attitude of the main structural plane is 0.38, the coefficient of influence of the attitude of the main structural plane is 0.14, the slope angle of the step is approximately 85°, and the step height is H. B It is approximately 10m.

[0083] Step S502: Determine the hole depth H and the appropriate fracturing device model based on the step height, and calculate the gas explosion energy λ per meter according to formula (5);

[0084] In gas explosions, fracturing devices come in fixed models, each with a fixed length and blast energy. The hole depth H is generally an integer multiple of the fracturing device length plus the plugging length, and the hole depth H is less than or close to the step height H. B Therefore, the borehole depth at this site is set at 5 meters. A Φ89×4×1200 disposable fracturing device is used, with 3 fracturing devices placed in each borehole. The plugging length is 1.4m. Thus, the gas explosion energy λ per meter along the drilling direction is 548.7kJ / m.

[0085] Step S503: Input the parameters into the trained neural network model to predict the reasonable optimal resistance line W. O and reasonable optimal unit consumption e O ;

[0086] The point load strength of the rock block I s(50) Rock block impedance ρV pr Rock mass integrity index K vThe six parameters—K5 (the influence coefficient of the main structural plane attitude), α (the slope angle of the step), and λ (the energy of the gas explosion per meter)—are input into a trained BP neural network to obtain the reasonable resistance line W. O and reasonable unit consumption e O The values ​​are 1.48m and 200.8kJ / m, respectively. 3 .

[0087] Step S504: Determine the hole layout parameters;

[0088] The distance from the first row of holes to the edge of the step is the resistance line, defined as W. O That is, 1.5m. Because gas blasting requires a high level of free face for rock breaking, it is generally carried out in actual blasting using a row-by-row or hole-by-hole initiation method. After the first row of holes is detonated, a free face is created for the second row. Therefore, the row spacing is also defined as b = W. O =1.5m, and the spacing between holes in the same row is calculated according to formula (7), which is approximately 1.8m.

Claims

1. A method for predicting gas explosion parameters and designing borehole layout based on backpropagation neural networks, characterized in that, Includes the following steps: Step S1: Obtain the physical and mechanical parameters of the rock mass in the blasting area, including: the point load strength I of the rock block. s(50) Wave impedance ρV of the rock block pr Rock mass integrity index K v The influence coefficient of the attitude of the main structural planes, K5; Step S2: Conduct multiple single-hole gas blasting tests in the same site, measure and record the rock breaking volume of each test, and determine the optimal resistance line and other key blasting parameters under the corresponding site conditions, including blasting energy, unit consumption and gas blasting energy per hole length. Step S3: Repeat steps S1 and S2 at blasting sites under different geological conditions, accumulate data through multiple experiments, and statistically analyze the relevant data and key blasting parameters obtained from the experiments to establish a database of gas blasting rock breaking operations that includes different geological conditions. Step S4: Construct and train a gas explosion parameter prediction model based on a backpropagation neural network. Input training data includes physical and mechanical parameters and experimental data. Output predicted gas explosion parameters. Optimize the neural network structure through the backpropagation algorithm to improve prediction accuracy. Step S5: Based on the neural network model trained in step S4, use the model to optimize the hole layout parameters for gas explosion, including the reasonable setting of resistance line, hole spacing, row spacing, and hole depth. Step S2 includes the following steps: Step S201: Conduct multiple single-hole gas blasting rock-breaking tests at the same site, keeping the test conditions consistent each time, adjusting only the resistance line W. Determine the maximum rock-breaking volume V by statistically analyzing the rock-breaking volume V of each test. O The corresponding optimal resistance line W under these site conditions O Step S202: Calculate the explosion energy E of the single-hole test according to the explosion energy calculation formula for compressed gas and water vapor containers. g : In the formula, k is the adiabatic index of carbon dioxide; p is the burst pressure; p atm It is atmospheric pressure; V0 is the initial volume of gas in a single explosion; Step S203: Based on the blast energy E g Calculate the optimal unit consumption e under the corresponding rock mass properties and site conditions. O : And the gas explosion energy λ per meter along the borehole: In the formula, H is the hole depth and T is the plugging length.

2. The method for predicting gas explosion parameters and designing aperture layout based on backpropagation neural network as described in claim 1, characterized in that, In step S1, ρ is the density of the rock block, and K is the rock mass integrity index. v Calculate according to formula (1): In the formula, V pm It is the elastic longitudinal wave velocity of the rock mass, V pr It is the elastic longitudinal wave velocity of the rock block; the influence coefficient K5 of the main structural plane attitude is calculated according to formula (2): In the formula, F1 represents the influence of the angle between the structural surface dip and the broken surface dip, F2 represents the influence of the structural surface dip angle, and F3 represents the influence of the difference between the structural surface dip angle and the slope dip angle.

3. The method for predicting gas explosion parameters and designing aperture layout based on backpropagation neural network as described in claim 1, characterized in that, Step S4 includes the following steps: Step S401: Based on the characteristics and correlations of the gas explosion parameters, determine the topology of the backpropagation neural network and construct a preliminary neural network model; Step S402: Select an activation function suitable for predicting gas explosion parameters and set initial weights; Step S403: Perform normalization preprocessing on the sample data, including normalization of the input samples and outlier removal; Step S404: Using the normalized sample data, train the neural network using the backpropagation algorithm and optimize the weights through multiple iterations to form a neural network model for predicting gas explosion parameters.

4. The method for predicting gas explosion parameters and designing aperture layout based on backpropagation neural network as described in claim 3, characterized in that, In step S401, the preliminary neural network model includes an input layer, a hidden layer, and an output layer, wherein the hidden layer network structure is designed as follows: Rock block point load strength I s(50) Rock block impedance ρV pr Rock mass integrity index K v The main structural plane attitude influence coefficient K5, step slope dip angle α, and gas explosion energy per meter λ are used as input indicators, corresponding to neurons in the input layer; with a reasonable optimal resistance line W O and reasonable optimal unit consumption e O The output metrics correspond to the neurons in the output layer; the number of neurons in the hidden layer is dynamically set according to the number of training samples.

5. The method for predicting gas explosion parameters and designing aperture layout based on backpropagation neural network as described in claim 3, characterized in that, In step S402, the activation function of the hidden layer neurons is a sigmoid-type logarithmic function logsig, and the activation function of the output layer neurons is a purelin-type linear function; the initial weights are generated by random numbers in the range of (-1,1).

6. The method for predicting gas explosion parameters and designing aperture layout based on backpropagation neural network as described in claim 3, characterized in that, In step S403, the original data is normalized according to the following method: In the formula, X represents the original data; X max X min These are the maximum and minimum values ​​of the original data; T represents the target data; T max T min These are the maximum and minimum values ​​of the target data.

7. The method for predicting gas explosion parameters and designing aperture layout based on backpropagation neural network as described in claim 3, characterized in that, In step S404, the neural network is trained using the feedforward multilayer neural network error backpropagation learning algorithm. The training function is the Bayesian regularized training function trainbr, and the network expected error, learning rate, and number of training iterations are set as training parameters. When the error is less than or equal to the set value, the network weights and threshold are saved, and the training ends.

8. The method for predicting gas explosion parameters and designing aperture layout based on backpropagation neural network as described in claim 1, characterized in that, Step S5 specifically includes the following steps: Step S501: Obtain rock mass parameters in the blasting area, including the point load strength I of rock blocks on the step slope. s(50) Rock block impedance ρV pr Rock mass integrity index K v The main structural plane attitude influence coefficient K5, the step slope dip angle α, and the step height H B ; Step S502: Determine the hole depth H and the appropriate fracturing device model based on the step height, and calculate the gas explosion energy λ per meter according to formula (5); Step S503: Input the parameters into the trained neural network model to predict the reasonable optimal resistance line W. O and reasonable optimal unit consumption e O ; Step S504: Determine the hole layout parameters and adopt a triangular hole layout, wherein the distance from the first row of holes to the edge of the step is W. O The spacing between holes in the same row is: In the formula, b is the row spacing, b = W O According to the design requirements, the holes are arranged in a triangular pattern.

Images