Coal face mine pressure prediction method and device and electronic equipment

Abstract

The invention provides a coal face mine pressure prediction method and device and electronic equipment, and the method comprises the steps: obtaining the working condition data of a target coal face, and building at least one data factor of the target coal face based on the working condition data; for any data factor, calculating an initial granulation parameter and a trend strength parameter of the data factor; adjusting the initial granulation parameter based on the trend strength parameter to generate a target granulation parameter; and constructing a target prediction data interval based on the target granulation parameter and the data factor, and inputting the target prediction data interval into a mine pressure prediction model to output predicted mine pressure data of the target coal face, the mine pressure prediction model comprising a space attention mechanism unit. According to the method, the change characteristics of the data can be more accurately captured by performing granularity self-adaptive adjustment on each data factor, and meanwhile, more reliable support is provided for mine pressure prediction by converting the data values of the characteristic factors into uncertainty capable of better presenting the mine pressure.

Description

Technical Field

[0001] The present invention relates to the field of coal mining safety technology, and in particular to a method, device, and electronic equipment for predicting mine pressure at a coal mining working face. Background Art

[0002] In the field of coal mining, the prediction of mine pressure at the coal mining face is crucial to safe production and efficient operation. As mining extends to deeper areas, the geological conditions become complex and the pattern of mine pressure is difficult to predict. Traditional mine pressure prediction methods, such as empirical analogy and mechanical analysis, rely on expert experience and simple mechanical models. In complex mining environments, the prediction accuracy and reliability are insufficient. Summary of the invention

[0003] The present disclosure aims to solve one of the technical problems in the related art at least to some extent.

[0004] To this end, one purpose of the present disclosure is to propose a method for predicting mine pressure in a coal mining face.

[0005] The second objective of the present disclosure is to provide a device for predicting mine pressure in a coal mining face.

[0006] A third objective of the present disclosure is to provide an electronic device.

[0007] A fourth object of the present disclosure is to provide a non-transitory computer-readable storage medium.

[0008] A fifth object of the present disclosure is to provide a computer program product.

[0009] To achieve the above-mentioned purpose, the first aspect of the present disclosure proposes a method for predicting mine pressure in a coal mining face, comprising: obtaining operating data of a target coal mining face, and establishing at least one data factor of the target coal mining face based on the operating data, wherein the data factor includes at least a spatial characteristic factor; for any data factor, calculating an initial granulation parameter and a trend intensity parameter of the data factor; adjusting the initial granulation parameter based on the trend intensity parameter to generate a target granulation parameter; constructing a target prediction data interval based on the target granulation parameter and the data factor, and inputting the target prediction data interval into a mine pressure prediction model to output predicted mine pressure data of the target coal mining face, wherein the mine pressure prediction model includes a spatial attention mechanism unit.

[0010] According to one embodiment of the present disclosure, constructing a target prediction data interval based on the target granulation parameter and the data factor includes: calculating a data lower limit value and a data upper limit value of the data factor based on the data values ​​of the target granulation parameter and the data factor; and constructing a data interval of the data factor based on the data lower limit value and the data upper limit value as the target prediction data interval.

[0011] According to one embodiment of the present disclosure, the calculation of the data lower limit value and the data upper limit value of the data factor based on the target granulation parameter and the data value of the data factor includes: taking the difference between the data value of the data factor and the target granulation parameter as the data lower limit value, and taking the sum of the data value of the data factor and the target granulation parameter as the data upper limit value.

[0012] According to one embodiment of the present disclosure, the adjusting the initial granulation parameters based on the trend strength parameters to generate target granulation parameters includes: comparing the trend strength parameters with a first trend strength threshold and a second trend strength threshold, respectively; and adjusting the initial granulation parameters based on the comparison result to generate the target granulation parameters.

[0013] According to one embodiment of the present disclosure, the adjusting the initial granulation parameter based on the comparison result to generate the target granulation parameter includes: obtaining an adjustment factor; in response to the trend strength parameter being less than the first trend strength threshold, adding 1 to the adjustment factor as a first target adjustment factor, and multiplying the first target adjustment factor by the initial granulation parameter to calculate and obtain the target granulation parameter; or, in response to the trend strength parameter being greater than or equal to the first trend strength threshold, and the trend strength parameter being less than or equal to the second trend strength threshold, using the initial granulation parameter as the target granulation parameter; or, in response to the trend strength parameter being greater than the second trend strength threshold, subtracting 1 from the adjustment factor as a second target adjustment factor, and multiplying the second target adjustment factor by the initial granulation parameter to calculate and obtain the target granulation parameter.

[0014] According to one embodiment of the present disclosure, before the initial granulation parameters are adjusted based on the comparison results to generate the target granulation parameters, the method further includes: obtaining window data corresponding to the adjustment coefficient and the data factor, and obtaining candidate granulation parameters after the initial granulation parameters are adjusted based on the comparison results; calculating the moving standard deviation of the data factor based on the window data, and calculating the fluctuation factor of the data factor based on the moving standard deviation and the adjustment coefficient; and calculating the target granulation parameters based on the fluctuation factor and the candidate granulation parameters.

[0015] According to one embodiment of the present disclosure, the calculating the initial granulation parameter and the trend strength parameter of the data factor includes: calculating the data factor based on a local weighted regression algorithm to calculate and obtain the initial granulation parameter and the trend strength parameter.

[0016] According to one embodiment of the present disclosure, the mine pressure prediction model is established, including: obtaining a plurality of candidate information particles of the target coal mining working face; calculating coverage value, specificity value, stability value and variance value based on the candidate information particles; calculating respective weight values ​​of the coverage value, the specificity value and the stability value based on the variance value; constructing the objective function of the target coal mining working face based on the weight value, the coverage value, the specificity value and the stability value, and establishing the mine pressure prediction model based on the objective function.

[0017] To achieve the above-mentioned purpose, the second aspect embodiment of the present disclosure proposes a coal mining face mine pressure prediction device, including: an acquisition module, used to acquire the operating data of the target coal mining face, and establish at least one data factor of the target coal mining face based on the operating data, and the data factor includes at least a spatial characteristic factor; a calculation module, used to calculate the initial granulation parameter and trend intensity parameter of the data factor for any data factor; an adjustment module, used to adjust the initial granulation parameter based on the trend intensity parameter to generate a target granulation parameter; a prediction module, used to construct a target prediction data interval based on the target granulation parameter and the data factor, and input the target prediction data interval into a mine pressure prediction model to output the predicted mine pressure data of the target coal mining face, and the mine pressure prediction model includes a spatial attention mechanism unit.

[0018] To achieve the above-mentioned purpose, the third aspect embodiment of the present disclosure proposes an electronic device, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to implement the method for predicting mining pressure in a coal mining face as described in the first aspect embodiment of the present disclosure.

[0019] To achieve the above-mentioned purpose, the fourth aspect embodiment of the present disclosure proposes a non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to implement the coal mining face mine pressure prediction method as described in the first aspect embodiment of the present disclosure.

[0020] To achieve the above-mentioned purpose, the fifth aspect of the present disclosure proposes a computer program product, including a computer program, which, when executed by a processor, is used to implement the method for predicting mine pressure in a coal mining face as described in the first aspect of the present disclosure.

[0021] Therefore, by adaptively adjusting the granularity of each data factor, the changing characteristics of the data can be captured more accurately. At the same time, by converting the data values ​​of the characteristic factors into a data that can better present the uncertainty of the mine pressure, more reliable support is provided for the mine pressure prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 It is a schematic diagram of a method for predicting the mine pressure in a coal mining face according to an embodiment of the present disclosure; Figure 2 It is a schematic diagram of another method for predicting the mine pressure of a coal mining face according to an embodiment of the present disclosure; Figure 3 It is a schematic diagram of another method for predicting the mine pressure of a coal mining face according to an embodiment of the present disclosure; Figure 4 is a schematic diagram of an optimization process using IGA according to one embodiment of the present disclosure; Figure 5 It is a schematic diagram of a device for predicting mine pressure in a coal mining face according to one embodiment of the present disclosure; Figure 6 is a schematic diagram of an electronic device according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0023] Embodiments of the present disclosure are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present disclosure, and should not be construed as limiting the present disclosure.

[0024] The acquisition, storage, use, and processing of data in the technical solution disclosed in this disclosure are in compliance with the relevant provisions of relevant laws and regulations.

[0025] It should be noted that in the embodiments of the present application, certain software, components, models and other existing solutions in the industry may be mentioned, and they should be considered as exemplary. Their purpose is only to illustrate the feasibility of implementing the technical solution of the present application, but it does not mean that the applicant has or will necessarily use the solution.

[0026] Figure 1 Schematic diagram of a method for predicting the mine pressure in a coal mining face according to an embodiment of the present disclosure. Figure 1 As shown, the method for predicting the mine pressure of a coal mining face includes the following steps: S101, obtaining working condition data of a target coal mining working face, and establishing at least one data factor of the target coal mining working face based on the working condition data, wherein the data factor includes at least a spatial characteristic factor.

[0027] The method for predicting the mine pressure of a coal mining face of an embodiment of the present application can be applied to the scenario of predicting the mine pressure of a coal mining face. The executor of the mine pressure prediction of a coal mining face of an embodiment of the present application can be the device for predicting the mine pressure of a coal mining face of an embodiment of the present application, and the device for predicting the mine pressure of a coal mining face can be set on an electronic device.

[0028] In current technologies, models (such as Long Short-Term Memory (LSTM)) only focus on time series features and ignore the spatial correlation between sensors. In this disclosure, by adding spatial feature factors to the working condition data and adding local spatial attention mechanism units to the mine pressure prediction model, it is possible to predict the mine pressure of the target coal mining face in both time and space dimensions.

[0029] It should be noted that, in addition to the spatial characteristic factor, the data factor may also include multiple factors, which are not limited here, for example, time characteristic factors, pressure characteristic factors, temperature characteristic factors, etc.

[0030] In the disclosed embodiment, there are many methods for obtaining the working condition data of the target coal mining face, which are not limited here. For example, the data can be collected and obtained through a sensor detection system, and the working condition data can be collected and analyzed through an automatic control system, and the data can be simulated and analyzed through data analysis and simulation software.

[0031] S102: For any data factor, calculate the initial granulation parameter and trend strength parameter of the data factor.

[0032] In current technology, with the rise of intelligent prediction technology, prediction systems and methods based on granular neural networks have attracted attention due to their unique data processing advantages. They can divide data into different granularities and mine data features. However, existing information granularity allocation schemes have limitations. On the one hand, they are mostly based on fixed rules or simple statistical analysis to determine information granularity, which is difficult to dynamically adjust with the real-time changes in mine pressure data. Because the mine pressure of the coal mining face is affected by many dynamic factors such as changes in coal seam thickness, differences in roof lithology, and mining progress, information granularity needs to be adaptively adjusted to reflect data characteristics; on the other hand, existing schemes are not flexible enough in balancing data generalization and detail. If the granularity is too coarse, key details will be lost, resulting in a decrease in prediction accuracy. If the granularity is too fine, too much noise will be introduced, increasing model complexity and computational burden, and reducing prediction efficiency.

[0033] Information granularity refers to the degree to which data is divided into "grains" or units of different sizes. Finer information granularity means more detailed data representation, while coarser information granularity means a higher level of data abstraction.

[0034] Adaptive adjustment of information granularity is an intelligent data processing technology that aims to dynamically adjust the granularity of data (i.e., the degree of refinement or aggregation level of data) according to the characteristics of the data to optimize the effect of data analysis. This method is particularly suitable for scenarios that require processing large and complex data sets, such as mine pressure monitoring and underground working condition analysis. By adjusting the information granularity, the accuracy and efficiency of the model can be improved while reducing the consumption of computing resources.

[0035] The initial granulation parameter refers to the setting used to define the degree of refinement or aggregation level of the "granules" into which the data is initially divided during the adaptive adjustment of the information granularity.

[0036] The trend strength parameter is an indicator used to measure the significance of the change trend in each local area of ​​the data series. It is defined based on the local weighted mean and local slope calculated by local weighted regression and is used to guide the initial adjustment of granulation parameters.

[0037] In the embodiments of the present disclosure, there may be multiple methods for calculating the initial granulation parameters and trend strength parameters of the data factors, which are not limited here.

[0038] In one possible implementation, the data value of the preset algorithm data factor can be used for calculation to obtain the initial granulation parameter and trend strength parameter of the data factor.

[0039] In another possible implementation, the data value of the data factor can also be processed by a processing model to output the initial granulation parameter and trend strength parameter of the data factor. The processing model is trained in advance and can be stored in the storage space of the electronic device to facilitate retrieval and use when needed.

[0040] S103, adjusting the initial granulation parameters based on the trend strength parameters to generate target granulation parameters.

[0041] In the current technology, there are many ways to adjust the granulation parameters. For example, the adjustment methods in the current technology can be as follows C1 and C2. Information granularity symmetric allocation scheme C1 Each time series granularity parameter is symmetrically spaced, that is, εi-=εi+=εi.

[0042]

[0043] Where i = 1, 2, …, n Information granularity non-uniform distribution scheme C2 Different time series data have different levels of granularity.

[0044] Where i = 1, 2, …, n The solution in the present disclosure is to perform an adaptive weighted granularity allocation scheme on the initial granulation parameters by using the trend strength parameter.

[0045] S104, constructing a target prediction data interval based on the target granulation parameter and the data factor, and inputting the target prediction data interval into the mine pressure prediction model to output the predicted mine pressure data of the target coal mining working face, the mine pressure prediction model includes a spatial attention mechanism unit.

[0046] In the disclosed embodiment, the working condition data of the target coal mining face is first obtained, and at least one data factor of the target coal mining face is established based on the working condition data, the data factor at least including the spatial characteristic factor, and then for any data factor, the initial granulation parameter and the trend intensity parameter of the data factor are calculated, and then the initial granulation parameter is adjusted based on the trend intensity parameter to generate the target granulation parameter, and finally the target prediction data interval is constructed based on the target granulation parameter and the data factor, and the target prediction data interval is input into the mine pressure prediction model to output the predicted mine pressure data of the target coal mining face. Therefore, by adaptively adjusting the granularity of each data factor, the change characteristics of the data can be captured more accurately, and by converting the data value of the characteristic factor into a better representation of the uncertainty of the mine pressure, more reliable support is provided for the mine pressure prediction.

[0047] In the embodiment of the present disclosure, the initial granulation parameters and the trend strength parameters of the data factors are calculated. The data factors may be calculated based on a local weighted regression algorithm to obtain the initial granulation parameters and the trend strength parameters.

[0048] In one possible implementation, for each data point in the sequence , using Locally Weighted Regression (LWR) to calculate the local weighted average and the local slope In LWR calculation, given bandwidth h, data point right Weight Determined by the Gaussian kernel function, the formula is

[0049] Among them, the kernel function , which determines the different data points The degree of influence, distance The closer the point, the higher the weight. Local weighted average It is obtained by weighted summing and normalizing the surrounding data points. The calculation formula is:

[0050] Local slope Reflects the data points The change trend near the local weighted regression curve Derived with respect to i, we get The expression of is more complicated and is usually derived with the help of numerical methods (such as the finite difference method), for example, using the central difference formula:

[0051] in is a smaller increment, and its value needs to be reasonably determined based on the sampling frequency and change characteristics of the data. is the trend strength parameter of the i-th data factor.

[0052] In the above embodiment, the target prediction data interval is constructed based on the target granulation parameter and the data factor, and the target prediction data interval can also be constructed by Figure 2 Explaining further, the method includes: S201, calculating a data lower limit value and a data upper limit value of a data factor based on a target granulation parameter and a data value of the data factor.

[0053] In the embodiment of the present disclosure, after the target granulation parameter is obtained, the data value of the target granulation parameter and the data factor can be calculated by a preset algorithm to obtain the data lower limit value and the data upper limit value.

[0054] For example, the preset algorithm may be as follows:

[0055]

[0056] Among them, a is the lower limit of the data, b is the upper limit of the data, is the data value of the ith data factor, is the target granulation parameter of the i-th data factor.

[0057] S202, constructing a data interval of the data factor based on the data lower limit value and the data upper limit value as the target prediction data interval.

[0058] In the embodiment of the present disclosure, taking S201 as an example, the data interval is .

[0059] In the embodiment of the present disclosure, the data lower limit value and the data upper limit value of the data factor are first calculated based on the target granulation parameter and the data value of the data factor, and then the data interval of the data factor is constructed based on the data lower limit value and the data upper limit value as the target prediction data interval.

[0060] In the above embodiment, the data lower limit value and the data upper limit value of the data factor are first calculated based on the target granulation parameter and the data value of the data factor, and then the data interval of the data factor is constructed based on the data lower limit value and the data upper limit value as the target prediction data interval. Therefore, by establishing the target prediction data interval, not only can the accuracy and reliability of the mine pressure prediction be significantly improved, but also the resource allocation can be effectively optimized and the decision support capability can be enhanced. In addition, by defining the data range and constructing the data interval, the robustness and computational efficiency of the model are further improved.

[0061] In the above embodiment, the initial granulation parameters are adjusted based on the trend strength parameters to generate the target granulation parameters. Figure 3 Explaining further, the method includes: S301, comparing the trend strength parameter with a first trend strength threshold and a second trend strength threshold respectively.

[0062] In the disclosed embodiment, the first trend strength threshold and the second trend strength threshold are critical values ​​for determining whether the initial granulation parameters need to be changed.

[0063] It should be noted that the first trend strength threshold and the second trend strength threshold are designed in advance and can be changed according to actual design needs, and no limitation is made here.

[0064] S302: Adjust the initial granulation parameters based on the comparison result to generate target granulation parameters.

[0065] In an embodiment of the present disclosure, an adjustment factor may be first obtained, and in response to a trend strength parameter being less than a first trend strength threshold, the sum obtained by adding 1 to the adjustment factor is used as a first target adjustment factor, and the first target adjustment factor is multiplied by the initial granulation parameter to calculate the target granulation parameter, or in response to a trend strength parameter being greater than or equal to a first trend strength threshold and the trend strength parameter being less than or equal to a second trend strength threshold, the initial granulation parameter is used as the target granulation parameter, or in response to a trend strength parameter being greater than a second trend strength threshold, the difference obtained by subtracting 1 from the adjustment factor is used as a second target adjustment factor, and the second target adjustment factor is multiplied by the initial granulation parameter to calculate the target granulation parameter.

[0066] In one possible implementation, define the trend strength parameter To measure the local trend strength at the data point. Assume that the initial granulation parameter is , the adjustment factor is According to the size of τ, the granulation parameters are preliminarily adjusted according to the following rules: When , it indicates that the data changes relatively slowly. , appropriately increase the granulation parameters to include more data information; when When , reducing the granulation parameter makes the information granules more focused on local details; when When the granulation parameters are kept unchanged, that is, Here and It is a pre-set trend strength threshold, which needs to be determined based on the historical fluctuation range of mine pressure data and actual experience.

[0067] In another possible implementation, in order to further consider the fluctuation characteristics of the data, a data fluctuation factor is introduced. The initial granulation parameter can be adjusted based on the comparison result first, and before the target granulation parameter is generated, the window data corresponding to the adjustment coefficient and the data factor can be obtained, and the candidate granulation parameter after the initial granulation parameter is adjusted based on the comparison result is obtained, and then the moving standard deviation of the data factor is calculated based on the window data, and the fluctuation factor of the data factor is calculated based on the moving standard deviation and the adjustment coefficient, and finally the target granulation parameter is calculated based on the fluctuation factor and the candidate granulation parameter.

[0068] In one possible implementation, the moving standard deviation of the mine pressure data can be calculated To measure the local fluctuation of the data. The window size is set to w, and the moving standard deviation The calculation formula is:

[0069] in is the average value of the data in the window Mean of the moving standard deviation for:

[0070] Finally, the granulation parameters after preliminary adjustment are adjusted according to the moving standard deviation. Make a secondary adjustment, the formula is

[0071] Where c is the adjustment coefficient. This formula enables the granulation parameters to be dynamically adjusted according to the real-time fluctuation of the data. In areas with large data fluctuations, the granulation parameters are further increased to better cover the data variation range; in areas with small fluctuations, the granulation parameters are maintained or appropriately reduced to improve the accuracy of the information granules. Through the above steps, this scheme can adaptively adjust the granularity of information granules according to the local trend and fluctuation characteristics of the mine pressure data, thereby better balancing the generalization and details of the data in the process of information granulation, and providing more representative input information for the subsequent mine pressure prediction based on granular neural network.

[0072] The inventors have found through experiments that the neural networks in current technologies (such as LSTM and Transformer) mostly adopt fixed granularity or static adjustment strategies, which are difficult to adapt to the dynamic changes of mine pressure data. The adaptive weighted granularity allocation scheme of this method: the trend strength of the data point is calculated by local weighted regression ( ) and volatility factor (moving standard deviation ), dynamically adjust granulation parameters For example, when hour, Reduce to focus on local details; when the data fluctuates violently, Further increase to cover the range of changes. Experiments show that the objective function value Q of the disclosed solution is improved by about 15% compared with the traditional fixed granularity solution, and the interval prediction coverage (cov) is stable at more than 74%, which is significantly better than the existing method.

[0073] In the disclosed embodiment, to establish the mine pressure prediction model, a plurality of candidate information particles of the target coal mining working face may be first obtained, and then the coverage value, specificity value, stability value and variance value may be calculated based on the candidate information particles, and then the weight values ​​of the coverage value, specificity value and stability value may be calculated based on the variance value, and finally the objective function of the target coal mining working face may be constructed based on the weight value, coverage value, specificity value and stability value, and the mine pressure prediction model may be established based on the objective function.

[0074] It should be noted that coverage measures the extent to which information granules can cover the initial data. It is expected that as much initial data as possible is included in the information granules. The higher the coverage, the better the information granules can cover the original data set and reduce data loss.

[0075] Specificity requires that information granules have clear meanings and express the meaning of the original data as specifically as possible. The higher the specificity, the more specific the expression of the information granules is, and the more accurately it can reflect the characteristics of the original data.

[0076] Stability measures the consistency of the data within an information granule. A stable information granule means that its internal data changes less and has lower volatility. The higher the stability, the more consistent the data within the information granule and the smaller the data fluctuation.

[0077] In one possible implementation, the objective function can be as follows:

[0078] Among them, cov is coverage (coverage, cov): it is expected that as much initial data as possible is included in the information granule. By calculating the granularity output ,in,

[0079] is a candidate information particle, It is the collection of information granule outputs.

[0080] Sp stands for specificity (sp): the information particles constructed are required to have clear meanings, that is, to express the meaning of the original data as specifically as possible. The calculation formula is:

[0081] and are the maximum and minimum values ​​of the kth information particle respectively.

[0082] Sta is stability (sta): for information particles ,The stability is measured by calculating the ratio of the range to the mean of the data in the information granule, that is, , overall stability

[0083] in, and are the maximum and minimum values ​​of the i-th information particle respectively.

[0084] Var is the prediction variance (var): Assuming that the prediction result set is obtained under different granulation parameters , the prediction variance is , in is the average value of the prediction results.

[0085] Through this objective function, the performance of information particles in different aspects can be comprehensively considered, and the weights can be adjusted according to actual needs. , optimize the information granulation effect and lay the foundation for subsequent analysis and prediction.

[0086] The inventors found through experiments that the current methods usually use mean square error (MSE) as the only optimization target, which leads to the inability to balance generalization and detail when granulating information. This method proposes a multi-index fusion objective function:

[0087] Through weighted optimization of coverage (cov), specificity (sp), stability (sta) and prediction variance (var), it is ensured that the information particles contain sufficient data (high cov) and avoid redundancy (high sp). In the optimization process, an improved genetic algorithm (IGA) is used, such as Figure 4As shown in Figure 1, it is based on the traditional genetic algorithm and adds an elite retention strategy and an adaptive crossover and mutation probability mechanism. Assume that the individual fitness is , crossover probability and mutation probability According to the formula, the adaptive adjustment is hour, ,when hour, ,in is the maximum fitness of the current population, is the average fitness, is the current individual fitness, The parameters are pre-set. With the help of information granule optimization configuration, objective function improvement and this improved genetic algorithm, the mine pressure data information granulation can be efficiently processed, and the optimal granulation parameters can be found by optimizing the objective function to improve the performance of the mine pressure prediction model, which will help the safe production of coal mines. In actual applications, parameters need to be adjusted according to the characteristics of mine pressure data and prediction needs, and the framework is scalable and can be optimized and expanded according to different situations.

[0088] Spatial attention mechanism processing: Traditional models (such as Long Short-Term Memory (LSTM)) only focus on time series features and ignore the spatial correlation between sensors. This method achieves two breakthroughs through the spatiotemporal attention mechanism: Local spatial attention mechanism unit: Focus on the correlation of different time series collected by a single sensor, calculate the attention weight, determine the importance of local features in prediction, and obtain the local spatial attention output vector. For each sensor's time series data, analyze the intrinsic connection between local time series according to data characteristics and spatial position, and provide local spatial feature information for subsequent predictions.

[0089] Global spatial attention mechanism unit: Considering the dynamic impact of other sensor target sequences on specific sensors, the attention weights between sensors are calculated, and the relevant sensors are adaptively selected to participate in the prediction based on the geographic spatial similarity. After updating the attention weights, the global spatial attention output vector is calculated. This process will consider the mutual influence between sensors at different support positions, comprehensively consider the spatial relationship of the entire coal mining face, and improve the accuracy of the prediction.

[0090] The inventors found through experiments that after the introduction of spatiotemporal attention, the root mean square error (RMSE) of the model loss function dropped from 0.931 to 0.904, and the prediction curve was more consistent with the true value, proving that the effective use of spatial information significantly improved the prediction accuracy.

[0091] Corresponding to the coal mining face mine pressure prediction methods provided in the above-mentioned embodiments, an embodiment of the present disclosure further provides a coal mining face mine pressure prediction device. Since the coal mining face mine pressure prediction device provided in the embodiment of the present disclosure corresponds to the coal mining face mine pressure prediction methods provided in the above-mentioned embodiments, the implementation methods of the above-mentioned coal mining face mine pressure prediction methods are also applicable to the coal mining face mine pressure prediction device provided in the embodiment of the present disclosure, and will not be described in detail in the following embodiments.

[0092] Figure 5 Schematic diagram of a device for predicting coal mining pressure in a coal mining face according to an embodiment of the present disclosure. Figure 5 As shown, the coal mining face mine pressure prediction device 500 includes: an acquisition module 510, a calculation module 520, an adjustment module 530 and a prediction module 540.

[0093] The acquisition module 510 is used to acquire the working condition data of the target coal mining working face, and establish at least one data factor of the target coal mining working face based on the working condition data, and the data factor at least includes a spatial characteristic factor.

[0094] The calculation module 520 is used to calculate the initial granulation parameter and trend strength parameter of any data factor.

[0095] The adjustment module 530 is used to adjust the initial granulation parameters based on the trend strength parameters to generate target granulation parameters.

[0096] The prediction module 540 is used to construct a target prediction data interval based on the target granulation parameters and data factors, and input the target prediction data interval into the mine pressure prediction model to output the predicted mine pressure data of the target coal mining working face. The mine pressure prediction model includes a spatial attention mechanism unit.

[0097] According to one embodiment of the present disclosure, a target prediction data interval is constructed based on target granulation parameters and data factors, including: calculating a data lower limit value and a data upper limit value of the data factor based on the data values ​​of the target granulation parameters and the data factors; and constructing a data interval of the data factor based on the data lower limit value and the data upper limit value as the target prediction data interval.

[0098] According to one embodiment of the present disclosure, a data lower limit value and a data upper limit value of a data factor are calculated based on a target granulation parameter and a data value of the data factor, including: taking the difference between the data value of the data factor and the target granulation parameter as the data lower limit value, and taking the sum of the data value of the data factor and the target granulation parameter as the data upper limit value.

[0099] According to one embodiment of the present disclosure, initial granulation parameters are adjusted based on trend strength parameters to generate target granulation parameters, including: comparing the trend strength parameters with a first trend strength threshold and a second trend strength threshold, respectively; and adjusting the initial granulation parameters based on the comparison result to generate the target granulation parameters.

[0100] According to one embodiment of the present disclosure, the initial granulation parameters are adjusted based on the comparison result to generate the target granulation parameters, including: obtaining an adjustment factor; in response to the trend strength parameter being less than a first trend strength threshold, adding 1 to the adjustment factor to obtain a first target adjustment factor, and multiplying the first target adjustment factor by the initial granulation parameter to calculate the target granulation parameter; or, in response to the trend strength parameter being greater than or equal to the first trend strength threshold, and the trend strength parameter being less than or equal to the second trend strength threshold, using the initial granulation parameter as the target granulation parameter; or, in response to the trend strength parameter being greater than the second trend strength threshold, subtracting 1 from the adjustment factor to obtain a second target adjustment factor, and multiplying the second target adjustment factor by the initial granulation parameter to calculate the target granulation parameter.

[0101] According to one embodiment of the present disclosure, before adjusting the initial granulation parameters based on the comparison results to generate the target granulation parameters, it also includes: obtaining window data corresponding to the adjustment coefficient and the data factor, and obtaining candidate granulation parameters after adjusting the initial granulation parameters based on the comparison results; calculating the moving standard deviation of the data factor based on the window data, and calculating the fluctuation factor of the data factor based on the moving standard deviation and the adjustment coefficient; calculating the target granulation parameters based on the fluctuation factor and the candidate granulation parameters.

[0102] According to one embodiment of the present disclosure, calculating the initial granulation parameters and trend strength parameters of the data factors includes: calculating the data factors based on a local weighted regression algorithm to obtain the initial granulation parameters and trend strength parameters.

[0103] Therefore, by adaptively adjusting the granularity of each data factor, the changing characteristics of the data can be captured more accurately. At the same time, by converting the data values ​​of the characteristic factors into a data that can better present the uncertainty of the mine pressure, more reliable support is provided for the mine pressure prediction.

[0104] In order to implement the above embodiment, the present disclosure further provides an electronic device 600, Figure 6 is a schematic diagram of an electronic device according to an embodiment of the present disclosure, such as Figure 6 As shown, the electronic device 600 includes: a processor 601 and a memory 602 that is communicatively connected to the processor, the memory 602 stores instructions that can be executed by at least one processor, and the instructions are executed by at least one processor 601 to implement the present disclosure. Figure 1-Figure 4 A method for predicting mine pressure in a coal mining face according to an embodiment.

[0105] In order to implement the above embodiments, the present disclosure also provides a non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to enable the computer to implement the above embodiments. Figure 1-Figure 4 A method for predicting mine pressure in a coal mining face according to an embodiment.

[0106] In order to implement the above embodiments, the present disclosure also provides a computer program product, including a computer program. When the computer program is executed by a processor, the computer program implements the above embodiments. Figure 1-Figure 4 A method for predicting mine pressure in a coal mining face according to an embodiment.

[0107] It should be noted that personal information from users should be collected for legitimate and reasonable purposes and should not be shared or sold outside of these legitimate uses. In addition, such collection / sharing should be carried out after receiving the user's informed consent, including but not limited to notifying the user to read the user agreement / user notice and sign the agreement / authorization including authorization of relevant user information before the user uses the function. In addition, any necessary steps should be taken to protect and safeguard access to such personal information data and ensure that others who have access to personal information data comply with its privacy policy and procedures.

[0108] The present application is expected to provide an implementation scheme for users to selectively block the use or access of personal information data. That is, the present disclosure is expected to provide hardware and / or software to prevent or block access to such personal information data. Once the personal information data is no longer needed, the risk can be minimized by limiting data collection and deleting the data. In addition, when applicable, such personal information is de-identified to protect the privacy of the user.

[0109] In the description of the aforementioned embodiments, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, unless they contradict each other.

[0110] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In the description of this application, the meaning of "plurality" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

[0111] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, fragment or portion of code comprising one or more executable instructions for implementing the steps of a custom logical function or process, and the scope of the preferred embodiments of the present application includes alternative implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by technicians in the technical field to which the embodiments of the present application belong.

[0112] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in combination with these instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that contains, stores, communicates, propagates or transmits a program for use by an instruction execution system, device or apparatus, or in combination with these instruction execution systems, devices or apparatuses. More specific examples (non-exhaustive list) of computer-readable media include the following: an electrical connection with one or more wires (electronic device), a portable computer disk box (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium and then editing, interpreting or processing in other suitable ways if necessary, and then stored in a computer memory.

[0113] It should be understood that the various parts of the present application can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0114] A person skilled in the art may understand that all or part of the steps in the method for implementing the above-mentioned embodiment may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, which, when executed, includes one or a combination of the steps of the method embodiment.

[0115] In addition, each functional unit in each embodiment of the present application may be integrated into a processing module, or each unit may exist physically separately, or two or more units may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0116] The storage medium mentioned above may be a read-only memory, a magnetic disk or an optical disk, etc. Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limiting the present application. A person of ordinary skill in the art may change, modify, replace and modify the above embodiments within the scope of the present application.

Claims

1. A method for predicting the mine pressure of a coal mining face, characterized in that: include: Acquire working condition data of a target coal mining working face, and establish at least one data factor of the target coal mining working face based on the working condition data, wherein the data factor includes at least a spatial characteristic factor; For any data factor, calculating the initial granulation parameter and trend strength parameter of the data factor; adjusting the initial granulation parameter based on the trend strength parameter to generate a target granulation parameter; A target prediction data interval is constructed based on the target granulation parameter and the data factor, and the target prediction data interval is input into a mine pressure prediction model to output predicted mine pressure data of the target coal mining face, wherein the mine pressure prediction model includes a spatial attention mechanism unit.

2. The method according to claim 1, characterized in that The constructing a target prediction data interval based on the target granulation parameter and the data factor includes: Calculate a data lower limit value and a data upper limit value of the data factor based on the target granulation parameter and the data value of the data factor; A data interval of the data factor is constructed based on the data lower limit value and the data upper limit value as the target prediction data interval.

3. The method according to claim 2, characterized in that The calculating the data lower limit value and the data upper limit value of the data factor based on the target granulation parameter and the data value of the data factor comprises: The difference between the data value of the data factor and the target granulation parameter is used as the data lower limit value, and the sum of the data value of the data factor and the target granulation parameter is used as the data upper limit value.

4. The method according to claim 1, characterized in that: The adjusting the initial granulation parameter based on the trend strength parameter to generate a target granulation parameter includes: comparing the trend strength parameter with a first trend strength threshold and a second trend strength threshold, respectively; The initial granulation parameters are adjusted based on the comparison result to generate the target granulation parameters.

5. The method according to claim 4, characterized in that The adjusting the initial granulation parameters based on the comparison result to generate the target granulation parameters includes: Get the adjustment factor; In response to the trend strength parameter being less than the first trend strength threshold, adding 1 to the adjustment factor to obtain a first target adjustment factor, and multiplying the first target adjustment factor by the initial granulation parameter to calculate the target granulation parameter; or In response to the trend strength parameter being greater than or equal to the first trend strength threshold, and the trend strength parameter being less than or equal to the second trend strength threshold, taking the initial granulation parameter as the target granulation parameter; or, In response to the trend strength parameter being greater than the second trend strength threshold, a difference obtained by subtracting 1 from the adjustment factor is used as a second target adjustment factor, and the second target adjustment factor is multiplied by the initial granulation parameter to calculate the target granulation parameter.

6. The method according to claim 4 or 5, characterized in that: Before the initial granulation parameters are adjusted based on the comparison result to generate the target granulation parameters, the method further includes: Acquire window data corresponding to the adjustment coefficient and the data factor, and acquire candidate granulation parameters after adjusting the initial granulation parameters based on the comparison result; Calculating a moving standard deviation of the data factor based on the window data, and calculating a fluctuation factor of the data factor based on the moving standard deviation and the adjustment coefficient; The target granulation parameter is calculated based on the fluctuation factor and the candidate granulation parameter.

7. The method according to claim 1, characterized in that The calculating of the initial granulation parameter and the trend strength parameter of the data factor comprises: The data factors are calculated based on a local weighted regression algorithm to obtain the initial granulation parameter and the trend strength parameter.

8. The method according to claim 1, characterized in that Establishing the mine pressure prediction model includes: Acquire a plurality of candidate information particles of the target coal mining working face; Calculate coverage value, specificity value, stability value and variance value based on the candidate information particles; Calculate the weight values ​​of the coverage value, the specificity value, and the stability value based on the variance value; The objective function of the target coal mining face is constructed based on the weight value, the coverage value, the specificity value, and the stability value, and the mine pressure prediction model is established based on the objective function.

9. A device for predicting coal mining working face pressure, characterized in that: include: An acquisition module, used for acquiring working condition data of a target coal mining working face, and establishing at least one data factor of the target coal mining working face based on the working condition data, wherein the data factor includes at least a spatial characteristic factor; A calculation module, used for calculating the initial granulation parameter and the trend strength parameter of any data factor; An adjustment module, configured to adjust the initial granulation parameter based on the trend strength parameter to generate a target granulation parameter; A prediction module is used to construct a target prediction data interval based on the target granulation parameter and the data factor, and input the target prediction data interval into a mine pressure prediction model to output the predicted mine pressure data of the target coal mining working face, wherein the mine pressure prediction model includes a spatial attention mechanism unit.

10. An electronic device, characterized in that: Including memory and processor; The processor runs a program corresponding to the executable program code by reading the executable program code stored in the memory, so as to implement the method according to any one of claims 1 to 8.

Images