All-working-condition comprehensive evaluation method for surrounding rock control effect of coal mine spraying support material

Abstract

The invention relates to the technical field of coal mine tunnel support, and provides a coal mine spraying support material surrounding rock control effect all-condition comprehensive evaluation method, which comprises the following steps of: determining a data feature vector according to to-be-evaluated spraying material test data under each condition; inputting the data feature vector into an all-working-condition comprehensive evaluation model, and outputting an all-working-condition control effect probability distribution vector by the all-working-condition comprehensive evaluation model; according to the all-working-condition control effect probability distribution vector, the all-working-condition control effect grade is determined, the purpose of conducting all-working-condition comprehensive evaluation on the coal mine spraying supporting material surrounding rock control effect by integrating multiple dimensions such as static load, dynamic load, repeated impact, loading and unloading circulation and dynamic and static load cooperation is achieved, and engineering practice is guided visually.

Description

Technical Field

[0001] This invention relates to the field of coal mine roadway support technology, and in particular to a comprehensive evaluation method for the control effect of coal mine sprayed support materials on surrounding rock under all working conditions. Background Technology

[0002] Coal resources are a crucial guarantee for my country's energy security. With the depletion of shallow resources, coal mining is gradually extending to deeper levels, leading to increasingly complex geological environments in deep tunnels. The surrounding rock is not only subjected to static loads of high ground stress, but also frequently faces dynamic impacts from mining activities, fault activation, or roof fractures, as well as cyclic loading and unloading effects due to alternating mining advances. This complex mechanical environment of "static load + dynamic load + fatigue + rheology" places extremely high demands on the performance of tunnel support materials.

[0003] Currently, the evaluation system for the rock control effect of sprayed support materials still has significant limitations, mainly in the following aspects: 1) Single evaluation dimension and lack of systematicity: Existing laboratory evaluation methods are usually for single working conditions. For example, simple static load testing cannot reflect the material's resistance to collapse under rockburst. Although there are test methods for single working conditions such as repeated impact, loading and unloading cycles, and dynamic-static combinations, these test results are often independent. It is difficult to judge the comprehensive applicability of the material through a single index. 2) Lack of quantitative assessment of "full working condition" adaptability: The failure of underground surrounding rock is often the result of the coupling of multiple mechanisms. One material may have extremely high static strength but insufficient toughness and cannot resist impact; another material may have good toughness but poor fatigue resistance and fail after multiple cyclic loading. Summary of the Invention

[0004] To address the problems existing in the prior art, this invention provides a comprehensive evaluation method for the rock control effect of coal mine sprayed support materials under all working conditions.

[0005] This invention provides a comprehensive evaluation method for the rock control effect of sprayed support materials in coal mines under all working conditions, including: Obtain test data of the coating material to be evaluated under various working conditions; The data feature vector is determined based on the test data of the sprayed material to be evaluated under each working condition; the data feature vector includes the test score for each working condition. The data feature vector is input into the comprehensive evaluation model for all working conditions, and the comprehensive evaluation model for all working conditions outputs the probability distribution vector of the control effect for all working conditions. The comprehensive evaluation model for all working conditions is a model for comprehensively evaluating the control effect of coal mine sprayed support materials on surrounding rock under all working conditions. It is obtained by training the sample data feature vector determined based on the sample sprayed material test data and the control effect level label corresponding to each working condition in the sample data feature vector as input through machine learning. The control effect level under all operating conditions is determined based on the probability distribution vector of the control effect under all operating conditions.

[0006] According to the present invention, a comprehensive evaluation method for the surrounding rock control effect of coal mine sprayed support materials under all working conditions is provided. The method further includes: obtaining a comprehensive evaluation model under all working conditions, including: Acquire sample spraying material test data under various working conditions, determine the sample data feature vector based on the sample spraying material test data under various working conditions, and determine the control effect level label corresponding to each working condition in the sample data feature vector; Based on the sample data feature vector, a virtual sample data feature vector is generated, and the control effect level label corresponding to each working condition in the virtual sample data feature vector is determined. The model is trained based on the feature vectors of sample data, the feature vectors of virtual sample data, and the control effect level labels corresponding to each working condition to construct a comprehensive evaluation model for all working conditions. The model parameters are constrained and updated based on a preset composite loss function, which includes a classification error loss function and a physical consistency constraint loss function.

[0007] According to the present invention, a comprehensive evaluation method for the rock control effect of sprayed support materials in coal mines under all working conditions is provided, wherein generating virtual sample data feature vectors based on the sample data feature vectors includes: The sample data feature vector is standardized to obtain the standardized sample data feature vector. The standardized sample data feature vectors are mapped to Gaussian distribution parameters of a pre-defined dimension latent space. Data sampling is performed in the latent space based on the reparameterization extraction method, and standard normal distribution noise is introduced into the sampled data to generate latent vectors. A decoder network is constructed to map the latent vectors onto the original feature space, generating virtual sample data feature vectors.

[0008] According to the present invention, a comprehensive evaluation method for the control effect of coal mine sprayed support material on surrounding rock under all working conditions is provided, wherein the Gaussian distribution parameters include the mean vector and the logarithmic variance vector.

[0009] According to the present invention, a comprehensive evaluation method for the control effect of coal mine sprayed support material on surrounding rock under all working conditions is provided, wherein the step of introducing standard normal distribution noise into the sampled data to generate a latent vector includes: Based on the mean vector, the log-variance vector, and the randomly introduced standard global distribution noise, the latent vector is determined using the following formula; in, For potential vectors, It is the mean vector. It is the variance vector. This is standard normally distributed noise.

[0010] According to the present invention, a comprehensive evaluation method for the surrounding rock control effect of coal mine sprayed support material is provided under all working conditions, including repeated impact, loading and unloading cycle, dynamic and static load coordination, static load foundation mechanics, and dynamic load impact resistance.

[0011] According to the comprehensive evaluation method for the control effect of coal mine sprayed support material in surrounding rock under all working conditions provided by the present invention, the standardization processing of the sample data feature vector to obtain the standardized sample data feature vector includes: The sample data feature vector is standardized using the following formula to obtain the standardized sample data feature vector; in, The standardized feature vector of the sample data. Let j be the test score for the j-th working condition. Let be the mean of all samples under the j-th working condition. Let be the standard deviation of all samples under the j-th working condition, where j∈{1,2,3,4,5}.

[0012] This invention also provides a comprehensive evaluation device for the rock control effect of coal mine sprayed support materials under all working conditions, comprising: The acquisition module is used to acquire test data of the spraying material to be evaluated under various working conditions; The determination module is used to determine the data feature vector based on the test data of the sprayed material to be evaluated under each working condition; the data feature vector includes the test score for each working condition; The processing module is used to input the data feature vector into the comprehensive evaluation model for all working conditions, and the comprehensive evaluation model for all working conditions outputs the probability distribution vector of the control effect for all working conditions; wherein, the comprehensive evaluation model for all working conditions is a model for comprehensively evaluating the control effect of coal mine sprayed support materials on surrounding rock under all working conditions, which is obtained by taking the sample data feature vector determined based on the sample sprayed material test data and the control effect level label corresponding to each working condition in the sample data feature vector as input and training it through machine learning; The evaluation module is used to determine the level of control effect under all operating conditions based on the probability distribution vector of the control effect under all operating conditions.

[0013] 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. When the processor executes the program, it implements a comprehensive evaluation method for the surrounding rock control effect of any of the above-described coal mine sprayed support materials.

[0014] The present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements a comprehensive evaluation method for the surrounding rock control effect of any of the above-mentioned coal mine sprayed support materials under all working conditions.

[0015] The present invention also provides a computer program product, including a computer program, which, when executed by a processor, implements a comprehensive evaluation method for the surrounding rock control effect of any of the above-mentioned coal mine sprayed support materials under all working conditions.

[0016] This invention provides a comprehensive evaluation method for the rock control effect of sprayed support materials in coal mines under all working conditions. The method determines data feature vectors based on test data of the sprayed material under various working conditions, inputs these feature vectors into a comprehensive evaluation model, and outputs a probability distribution vector of the control effect under all working conditions. Based on this probability distribution vector, the level of control effect under all working conditions is determined. This method achieves a comprehensive evaluation of the rock control effect of sprayed support materials in coal mines under all working conditions, considering multiple dimensions such as static load, dynamic load, repeated impact, loading and unloading cycles, and synergistic dynamic and static loads. This provides intuitive guidance for engineering practice. Attached Figure Description

[0017] 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.

[0018] Figure 1 This is a flowchart illustrating the comprehensive evaluation method for the control effect of coal mine sprayed support materials on surrounding rock under all working conditions provided by the present invention.

[0019] Figure 2 This is the overall architecture diagram of the comprehensive evaluation method for the surrounding rock control effect of coal mine sprayed support materials provided by the present invention.

[0020] Figure 3 This is a schematic diagram of the structure of the comprehensive evaluation device for the control effect of coal mine sprayed support material on surrounding rock under all working conditions provided by the present invention.

[0021] Figure 4 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

[0022] 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.

[0023] Figure 1 This invention provides a flowchart illustrating a comprehensive evaluation method for the surrounding rock control effect of sprayed support materials in coal mines under all working conditions. (See attached diagram.) Figure 1 The method includes the following steps: Step 11: Obtain test data of the coating material to be evaluated under various working conditions.

[0024] Step 12: Determine the data feature vector based on the test data of the sprayed material to be evaluated under each working condition; the data feature vector includes the test score for each working condition.

[0025] Step 13: Input the data feature vector into the full-condition comprehensive evaluation model, and the full-condition comprehensive evaluation model outputs the probability distribution vector of the control effect under all working conditions; wherein, the full-condition comprehensive evaluation model is a model for comprehensively evaluating the control effect of coal mine sprayed support materials on surrounding rock under all working conditions by taking the sample data feature vector determined based on the sample sprayed material test data and the control effect level label corresponding to each working condition in the sample data feature vector as input and training it through machine learning.

[0026] Step 14: Determine the level of control effect under all operating conditions based on the probability distribution vector of the control effect under all operating conditions.

[0027] Regarding steps 11-14, it should be noted that coal resources are a crucial guarantee of energy security. As shallow resources are depleted, coal mining is gradually extending to deeper levels, leading to increasingly complex geological environments in deep tunnels. The surrounding rock is not only subjected to static loads of high ground stress but also frequently faces dynamic impacts from mining activities, fault activation, or roof fractures, as well as cyclic loading and unloading effects due to alternating mining advances. This complex mechanical environment of "static load + dynamic load + fatigue + rheology" places extremely high demands on the performance of tunnel support materials.

[0028] Therefore, a comprehensive evaluation method is needed to assess the rock-surrounding control effect of sprayed support materials in coal mines across multiple dimensions, including static load, dynamic load, repeated impact, loading and unloading cycles, and repeated impact. In this context, various working conditions encountered in the geological environment of deep roadways can include repeated impact, loading and unloading cycles, dynamic and static load coordination, static load foundation mechanics, and dynamic load impact resistance.

[0029] In this invention, independent laboratory tests are first conducted on the coal mine sprayed support material under different working conditions to obtain test data for the sprayed material to be evaluated under each condition. Then, a data feature vector is determined based on the test data for each working condition; the data feature vector includes the test score for each working condition. The test scores for each working condition are: repeated impact adaptability score, loading and unloading cycle adaptability score, dynamic and static load synergistic adaptability score, static load foundation mechanical score, and dynamic load impact resistance score. Then, based on these test scores, a data feature vector corresponding to the test data of the sprayed material to be evaluated is constructed, i.e., data feature vector X = [x1, x2, x3, x4, x5]. T .

[0030] (1) Obtaining the repeated impact adaptability score (x1): Incremental cyclic impact tests with a split Hopkinson bar and triaxial cyclic impact tests were conducted. The number of impact cycles, dynamic failure strength, dynamic elastic modulus degradation rate, and fragment ejection kinetic energy were recorded. Based on the above indicators, a comprehensive score (S) reflecting the fatigue resistance and energy absorption durability of the rock under repeated dynamic loads was obtained. RI The overall score (S) RI () as a score for repeated shock adaptation.

[0031] (2) Obtaining the loading and unloading cycle adaptability score (x2): Single-axis cyclic loading and unloading tests, single-axis incremental cyclic loading and unloading tests, single-axis incremental stabilizing cyclic loading and unloading tests, and single-axis incremental stepped cyclic loading and unloading tests were conducted. Indicators such as the cumulative plastic strain index, the deterioration rate of the unloading elastic modulus, and the acoustic emission event rate were recorded. Based on the above indicators and the surrounding rock control effect evaluation model, a comprehensive score (S) characterizing the plastic accumulation control and stiffness maintenance ability of the sprayed rock under stress fluctuation environment was obtained. LU The overall score (S) LU () is used as a score for the adaptability of the loading and unloading cycle.

[0032] (3) Obtaining the dynamic and static load synergistic adaptability score (x3): Uniaxial compressive strength test, split Hopkinson bar test, and "bidirectional compression" test were conducted respectively. The average ejection velocity of the fragments, the average ejection momentum of the fragments, the average kinetic energy of the fragments, and the fragmentation degree coefficient were recorded. Based on the above indicators and the surrounding rock control effect evaluation model, a comprehensive score (S) characterizing the synergistic deformation ability of the sprayed rock under bidirectional compression and complex boundary conditions was obtained. CT The overall score (S) CT () is used as a score for the coordinated adaptability of dynamic and static loads.

[0033] (4) Obtaining the static load foundation mechanical score (x4): Conduct uniaxial compressive strength test, uniaxial tensile strength test, shear strength test data, point load strength test, and triaxial compressive strength test. Record the uniaxial compressive strength change rate, elastic modulus change rate, brittleness index change rate, and tensile reinforcement index, and obtain a comprehensive score (S) reflecting the strengthening and toughening effect of the material foundation. ST The overall score (S) ST () is the mechanical score for static load foundation.

[0034] (5) Obtaining dynamic load impact resistance score (x5): Conducting split Hopkinson bar tests and dynamic-static combined triaxial compression tests. Recording indicators such as the average ejection velocity, ejection momentum, and energy absorption efficiency index of rock fragments. Obtaining a comprehensive score (S) characterizing the impact resistance and dynamic fracture resistance of the sprayed rock under strong impact. DY The overall score (S) DY () represents the dynamic load impact resistance score.

[0035] In this invention, the data feature vector is input into the comprehensive evaluation model for all working conditions, and the comprehensive evaluation model for all working conditions outputs the probability distribution vector of the control effect under all working conditions. The comprehensive evaluation model for all working conditions is a model for comprehensively evaluating the control effect of coal mine sprayed support materials on surrounding rock under all working conditions. It is obtained by training through machine learning, using the sample data feature vector determined based on the sample sprayed material test data and the control effect level label corresponding to each working condition in the sample data feature vector as input.

[0036] Then, based on the probability distribution vector of the control effect under all operating conditions, the control effect level under all operating conditions is determined. It should be noted that the probability distribution vector of the control effect under all operating conditions represents the probability value of each control effect level. Based on the probability distribution vector of the control effect under all operating conditions, the final control effect level is determined as the most reasonable control effect level obtainable under all operating conditions.

[0037] The further method described above mainly explains the process of obtaining the comprehensive evaluation model for all working conditions, as follows: Acquire sample spraying material test data under various working conditions, determine the sample data feature vector based on the sample spraying material test data under various working conditions, and determine the control effect level label corresponding to each working condition in the sample data feature vector; Based on the feature vector of the sample data, a virtual sample data feature vector is generated, and the control effect level label corresponding to each working condition in the virtual sample data feature vector is determined; The model is trained based on the feature vectors of sample data, the feature vectors of virtual sample data, and the control effect level labels corresponding to each working condition to construct a comprehensive evaluation model for all working conditions. The model parameters are constrained and updated based on a preset composite loss function, which includes a classification error loss function and a physical consistency constraint loss function.

[0038] It should be noted that the number of physical samples in rock mechanics experiments is insufficient (small sample size). Therefore, it is necessary to expand the original physical samples. In this invention, sample spraying material test data under various working conditions are acquired. Based on the sample spraying material test data under various working conditions, sample data feature vectors are determined, and control effect level labels corresponding to each working condition are determined in the sample data feature vectors. Virtual sample data feature vectors are generated based on the sample data feature vectors, and control effect level labels corresponding to each working condition are determined in the virtual sample data feature vectors. At this point, the sample data feature vectors, virtual sample data feature vectors, and control effect level labels corresponding to each working condition constitute enhanced sample data. Based on this enhanced sample data, a nonlinear mapping model from multidimensional physical indicators to support effect evaluation levels is established using a Deep Fully Connected Network (DFCN), that is, a comprehensive evaluation model for all working conditions is constructed.

[0039] To ensure the model better aligns with the fundamental laws of rock mechanics, model parameters need optimization. In this invention, model parameters are constrained and updated based on a preset composite loss function, which includes a classification error loss function and a physical consistency constraint loss function. The classification error loss function measures the difference between the neural network's prediction and the "benchmark true value label," while the physical consistency constraint loss function ensures that the control effect level conforms to the fundamental laws of rock mechanics.

[0040] In a further step of the above method, the process of generating virtual sample data feature vectors based on the sample data feature vectors is explained in detail as follows: The feature vectors of the sample data are standardized to obtain standardized feature vectors of the sample data. The standardized sample data feature vectors are mapped to Gaussian distribution parameters of a pre-defined dimension latent space. Data sampling is performed in the latent space based on the reparameterization extraction method, and standard normal distribution noise is introduced into the sampled data to generate latent vectors. A decoder network is constructed to map the latent vectors onto the original feature space, generating virtual sample data feature vectors.

[0041] It should be noted that the encoder network receives the standardized real experimental feature vector x and maps it to Gaussian distribution parameters in a low-dimensional latent space, namely the mean vector μ and the log-variance vector logσ. 2 .

[0042] A reparameterization trick is employed for sampling in the latent space. Standard normal distribution noise ϵ∼N(0,I) is introduced to generate the latent vector z, calculated using the following formula: in, For potential vectors, It is the mean vector. It is the variance vector. The noise is a standard normal distribution. This processing method simulates the uncertainty and randomness of the internal microstructure of natural rock materials.

[0043] A decoder network is constructed to map the sampled latent vector z back to the original feature space, generating virtual sample data feature vectors x'. By interpolating and adding small perturbations within the latent space, a large number of virtual samples that conform to the physical distribution of the original data are generated. These virtual samples, together with the real samples, constitute the augmented training dataset D. train .

[0044] To eliminate the impact of data discrepancies on neural network weight updates, the Z-Score normalization method is used to process the original data.

[0045] The sample data feature vectors are standardized to obtain standardized sample data feature vectors, including: The sample data feature vector is standardized using the following formula to obtain the standardized sample data feature vector. in, The standardized feature vector of the sample data. Let j be the test score for the j-th working condition. Let be the mean of all samples under the j-th working condition. Let be the standard deviation of all samples under the j-th working condition, where j∈{1,2,3,4,5}.

[0046] In addition, a scene weighting factor (w) is introduced based on the actual geological conditions of the mines where the sprayed materials were applied. env As auxiliary input: When working in a mine prone to rock bursts: Increase S CT S DYand S RI The initial weights.

[0047] In soft rock mines with large deformation: Increase S ST and S LU The initial weights.

[0048] In this invention, during the model training process, a Deep FullyConnected Network (DFCN) is constructed to establish a nonlinear mapping model from multidimensional physical indicators to support effect evaluation levels.

[0049] 1) Network architecture design: Input layer: Receives the 5-dimensional feature vector generated in step 1.

[0050] Deep hidden layers: Set at least 4 fully connected layers, and arrange the number of neurons in an "inverted pyramid" structure (e.g., 64→32→16→8) to compress the feature dimension layer by layer and extract high-order abstract features.

[0051] Output layer: Contains 4 neurons, corresponding to four evaluation levels: I (excellent), II (good), III (medium), and IV (poor).

[0052] 2) Network optimization configuration: Activation function: The hidden layers uniformly adopt the Leaky ReLU activation function (α=0.01) to enhance the model's nonlinear expression ability in the negative range and prevent neuron inactivation.

[0053] Regularization mechanism: Dropout layers (with a dropout rate set to 0.3) and batch normalization layers are introduced between fully connected layers to prevent model overfitting and improve generalization ability.

[0054] Probability Output: The output layer uses the Softmax function, and the output sample represents the probability distribution vector p=[p I p II p III p IV ].

[0055] In this invention, a composite loss function incorporating physical constraints is defined for the physical consistency constraint loss function to guide the updating of neural network parameters and ensure that the evaluation results conform to the basic laws of rock mechanics.

[0056] a) Definition of loss function: Total loss function L total Classification error loss L class And physical consistency constraint loss L phy Weighted composition: Where λ is the weighting coefficient.

[0057] L class Part of it is used to measure the difference between the neural network's predictions and the "benchmark truth label".

[0058] Assume the current training batch size is N, and there are C = 4 evaluation levels (Level I, Level II, Level III, Level IV). L class Calculated using formula (25): y i,c : The true label of the i-th sample. For example, if sample i is level I, then yi=[1,0,0,0].

[0059] p i,c The predicted probability of the i-th sample calculated by the Softmax function from the output layer of the neural network.

[0060] b) Physical consistency constraints (L) phy This model is constructed based on the principle of monotonicity constraints. During training, sample pairs (A, B) are randomly selected. If sample A has better physical performance indicators (such as lower launch kinetic energy and higher integrity coefficient) than sample B, then the model's evaluation score (or level I probability) for sample A must be higher than that for sample B. If the model output violates this rule, a penalty loss is incurred. The calculation steps are as follows: Since the neural network outputs a probability vector, it is first converted into a continuous scalar score for numerical comparison. The weight vectors for each level are defined as w = [1.0, 0.7, 0.4, 0.1] (corresponding to levels I to IV). For the i-th sample, its expected predicted security score is calculated as follows: The higher the score, the better the model considers the material to be in terms of support effectiveness.

[0061] In the current training batch, M pairs of samples (i, j) are randomly selected. For each pair of samples, the physical performance relationship is calculated based on the data feature vector x.

[0062] Physical advantage determination: Calculate the weighted Euclidean distance between the input feature vectors of two samples or directly compare their "initial comprehensive scores". If the comprehensive physical index performance of sample i is significantly better than that of sample j (denoted as xi ≻ xj), then theoretically the model's prediction score must satisfy S safe (i) >S safe(j) .

[0063] Hinge Loss and its variants are used to penalize predictions that violate the aforementioned physical laws.

[0064] Ω: The set of sample pairs that satisfy the physical dominance relation xi>xj.

[0065] S safe (j) -S safe (i) If sample i has better physics, but the model gives sample j a higher score (i.e., Sj>Si), then the difference is positive, indicating an error.

[0066] ζ: Safety margin, usually set to 0.1. This means that the score of a good sample must not only be higher than that of a bad sample, but also at least 0.1 points higher, otherwise a small loss will occur.

[0067] : ReLU function. If the prediction order is correct (Si>Sj+ζ), the value in parentheses is negative, and the loss is zero; loss only occurs when the prediction order is incorrect or the discrimination is insufficient.

[0068] In this invention, the level to which a product belongs is determined based on the output probability distribution vector p: If max(p) = p I And p I >θ (θ is the confidence threshold, such as 0.8), is judged as Level I (excellent control effect); If max(p) = p II It was determined to be Level II (good control effect); If max(p) = p III It is classified as Level III (moderate control effect). If max(p) = p IV It is classified as Level IV (poor control effect).

[0069] See Figure 2 This is the overall architecture diagram of the comprehensive evaluation method for the surrounding rock control effect of coal mine sprayed support materials provided by the present invention.

[0070] The present invention provides a comprehensive evaluation method for the rock control effect of coal mine sprayed support materials under all working conditions. This method determines data feature vectors based on test data of the sprayed material under various working conditions, inputs these feature vectors into a comprehensive evaluation model, and outputs a probability distribution vector of the control effect under all working conditions. Based on this probability distribution vector, the level of control effect under all working conditions is determined. This method achieves a comprehensive evaluation of the rock control effect of coal mine sprayed support materials under all working conditions, considering multiple dimensions such as static load, dynamic load, repeated impact, loading and unloading cycles, and synergy between static and dynamic loads. This provides intuitive guidance for engineering practice.

[0071] The following describes the comprehensive evaluation device for the control effect of coal mine sprayed support material on surrounding rock under all working conditions provided by the present invention. The comprehensive evaluation device for the control effect of coal mine sprayed support material on surrounding rock under all working conditions described below can be referred to in correspondence with the comprehensive evaluation method for the control effect of coal mine sprayed support material on surrounding rock under all working conditions described above.

[0072] Figure 3 This diagram illustrates the structure of a comprehensive evaluation device for the surrounding rock control effect of coal mine sprayed support materials, provided by the present invention. (See attached diagram.) Figure 3 The device includes an acquisition module 31, a determination module 32, a processing module 33, and an evaluation module 34, wherein: The acquisition module is used to acquire test data of the spraying material to be evaluated under various working conditions; The determination module is used to determine the data feature vector based on the test data of the sprayed material to be evaluated under each working condition; the data feature vector includes the test score for each working condition. The processing module is used to input the data feature vector into the comprehensive evaluation model for all working conditions, and the comprehensive evaluation model for all working conditions outputs the probability distribution vector of the control effect for all working conditions. The comprehensive evaluation model for all working conditions is a model that is trained by machine learning, using the sample data feature vector determined based on the sample spraying material test data and the control effect level label corresponding to each working condition in the sample data feature vector as input. The evaluation module is used to determine the level of control effect under all operating conditions based on the probability distribution vector of the control effect under all operating conditions.

[0073] Since the apparatus of this embodiment is based on the same principle as the method of the above embodiment, more detailed explanations will not be repeated here.

[0074] It should be noted that, in the embodiments of the present invention, the relevant functional modules can be implemented by a hardware processor.

[0075] The present invention provides a comprehensive evaluation device for the control effect of coal mine sprayed support materials on surrounding rock under all working conditions. This device determines data feature vectors based on test data of the sprayed materials under various working conditions, inputs these feature vectors into a comprehensive evaluation model, and outputs a probability distribution vector of the control effect under all working conditions. Based on this probability distribution vector, the level of control effect under all working conditions is determined. This device achieves the goal of comprehensively evaluating the control effect of coal mine sprayed support materials on surrounding rock under all working conditions, considering multiple dimensions such as static load, dynamic load, repeated impact, loading and unloading cycles, and synergy between static and dynamic loads. This provides intuitive guidance for engineering practice.

[0076] Figure 4 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 4 As shown, the electronic device may include: a processor 41, a communication interface 42, a memory 43, and a communication bus 44. The processor 41, communication interface 42, and memory 43 communicate with each other via the communication bus 44. The processor 41 can call logical instructions from the memory 43 to execute a comprehensive evaluation method for the control effect of coal mine sprayed support materials on surrounding rock under all working conditions. This method includes: Acquire test data of the sprayed material to be evaluated under various working conditions; determine data feature vectors based on the test data of the sprayed material to be evaluated under various working conditions; the data feature vectors include the test scores for each working condition; input the data feature vectors into the comprehensive evaluation model for all working conditions, and output the probability distribution vector of the control effect under all working conditions from the comprehensive evaluation model for all working conditions; wherein, the comprehensive evaluation model for all working conditions is a model trained by machine learning, which takes the sample data feature vector determined based on the sample sprayed material test data and the control effect level labels corresponding to each working condition in the sample data feature vector as input, and is used for the comprehensive evaluation of the control effect of sprayed support materials in the surrounding rock under all working conditions; determine the control effect level under all working conditions based on the probability distribution vector of the control effect under all working conditions.

[0077] Furthermore, the logical instructions in the aforementioned memory 43 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, 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 described in 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.

[0078] 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 is able to execute a comprehensive evaluation method for the control effect of coal mine sprayed support materials on surrounding rock under all working conditions. The method includes: Acquire test data of the sprayed material to be evaluated under various working conditions; determine data feature vectors based on the test data of the sprayed material to be evaluated under various working conditions; the data feature vectors include the test scores for each working condition; input the data feature vectors into the comprehensive evaluation model for all working conditions, and output the probability distribution vector of the control effect under all working conditions from the comprehensive evaluation model for all working conditions; wherein, the comprehensive evaluation model for all working conditions is a model trained by machine learning, which takes the sample data feature vector determined based on the sample sprayed material test data and the control effect level labels corresponding to each working condition in the sample data feature vector as input, and is used for the comprehensive evaluation of the control effect of sprayed support materials in the surrounding rock under all working conditions; determine the control effect level under all working conditions based on the probability distribution vector of the control effect under all working conditions.

[0079] In another aspect, 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 a comprehensive evaluation method for the control effect of coal mine sprayed support materials on surrounding rock under all working conditions, the method comprising: Acquire test data of the sprayed material to be evaluated under various working conditions; determine data feature vectors based on the test data of the sprayed material to be evaluated under various working conditions; the data feature vectors include the test scores for each working condition; input the data feature vectors into the comprehensive evaluation model for all working conditions, and output the probability distribution vector of the control effect under all working conditions from the comprehensive evaluation model for all working conditions; wherein, the comprehensive evaluation model for all working conditions is a model trained by machine learning, which takes the sample data feature vector determined based on the sample sprayed material test data and the control effect level labels corresponding to each working condition in the sample data feature vector as input, and is used for the comprehensive evaluation of the control effect of sprayed support materials in the surrounding rock under all working conditions; determine the control effect level under all working conditions based on the probability distribution vector of the control effect under all working conditions.

[0080] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. 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.

[0081] 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., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0082] 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 comprehensive evaluation method for the rock control effect of sprayed support materials in coal mines under all working conditions, characterized in that, include: Obtain test data of the coating material to be evaluated under various working conditions; The data feature vector is determined based on the test data of the sprayed material to be evaluated under each working condition; the data feature vector includes the test score for each working condition. The data feature vector is input into the comprehensive evaluation model for all working conditions, and the comprehensive evaluation model for all working conditions outputs the probability distribution vector of the control effect for all working conditions. The comprehensive evaluation model for all working conditions is a model for comprehensively evaluating the control effect of coal mine sprayed support materials on surrounding rock under all working conditions. It is obtained by training the sample data feature vector determined based on the sample sprayed material test data and the control effect level label corresponding to each working condition in the sample data feature vector as input through machine learning. The control effect level under all operating conditions is determined based on the probability distribution vector of the control effect under all operating conditions.

2. The comprehensive evaluation method for the surrounding rock control effect of coal mine sprayed support materials according to claim 1, characterized in that, The method further includes: obtaining a comprehensive evaluation model for all operating conditions, including: Acquire sample spraying material test data under various working conditions, determine the sample data feature vector based on the sample spraying material test data under various working conditions, and determine the control effect level label corresponding to each working condition in the sample data feature vector; Based on the sample data feature vector, a virtual sample data feature vector is generated, and the control effect level label corresponding to each working condition in the virtual sample data feature vector is determined. The model is trained based on the feature vectors of sample data, the feature vectors of virtual sample data, and the control effect level labels corresponding to each working condition to construct a comprehensive evaluation model for all working conditions. The model parameters are constrained and updated based on a preset composite loss function, which includes a classification error loss function and a physical consistency constraint loss function.

3. The comprehensive evaluation method for the surrounding rock control effect of coal mine sprayed support materials according to claim 2, characterized in that, The step of generating a virtual sample data feature vector based on the sample data feature vector includes: The sample data feature vector is standardized to obtain the standardized sample data feature vector. The standardized sample data feature vectors are mapped to Gaussian distribution parameters of a pre-defined dimension latent space. Data sampling is performed in the latent space based on the reparameterization extraction method, and standard normal distribution noise is introduced into the sampled data to generate latent vectors. A decoder network is constructed to map the latent vectors onto the original feature space, generating virtual sample data feature vectors.

4. The comprehensive evaluation method for the surrounding rock control effect of coal mine sprayed support materials according to claim 3, characterized in that, The Gaussian distribution parameters include the mean vector and the log-variance vector.

5. The comprehensive evaluation method for the surrounding rock control effect of coal mine sprayed support materials according to claim 4, characterized in that, The process of introducing standard normal distribution noise into the sampled data to generate a latent vector includes: Based on the mean vector, the log-variance vector, and the randomly introduced standard global distribution noise, the latent vector is determined using the following formula; in, For potential vectors, It is the mean vector. It is the variance vector. This is standard normally distributed noise.

6. The comprehensive evaluation method for the surrounding rock control effect of coal mine sprayed support materials according to claim 1, characterized in that, The operating conditions include repeated impacts, loading and unloading cycles, dynamic and static load coordination, static load foundation mechanics, and dynamic load impact resistance.

7. The comprehensive evaluation method for the surrounding rock control effect of coal mine sprayed support materials according to claim 3, characterized in that, The standardization process for the sample data feature vector to obtain the standardized sample data feature vector includes: The sample data feature vector is standardized using the following formula to obtain the standardized sample data feature vector; in, The standardized feature vector of the sample data. Let j be the test score for the j-th working condition. Let be the mean of all samples under the j-th working condition. Let be the standard deviation of all samples under the j-th working condition, where j∈{1,2,3,4,5}.

8. A comprehensive evaluation device for the rock control effect of sprayed support materials in coal mines under all working conditions, characterized in that, include: The acquisition module is used to acquire test data of the spraying material to be evaluated under various working conditions; The determination module is used to determine the data feature vector based on the test data of the sprayed material to be evaluated under each working condition; the data feature vector includes the test score for each working condition; The processing module is used to input the data feature vector into the comprehensive evaluation model for all working conditions, and the comprehensive evaluation model for all working conditions outputs the probability distribution vector of the control effect for all working conditions; wherein, the comprehensive evaluation model for all working conditions is a model for comprehensively evaluating the control effect of coal mine sprayed support materials on surrounding rock under all working conditions, which is obtained by taking the sample data feature vector determined based on the sample sprayed material test data and the control effect level label corresponding to each working condition in the sample data feature vector as input and training it through machine learning; The evaluation module is used to determine the level of control effect under all operating conditions based on the probability distribution vector of the control effect under all operating conditions.

9. 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 program, it implements the comprehensive evaluation method for the surrounding rock control effect of coal mine sprayed support material as described in any one of claims 1-7.

10. 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 comprehensive evaluation method for the surrounding rock control effect of coal mine sprayed support material as described in any one of claims 1-7.

Images