An intelligent prediction method for the height of two zones in coal mine goaf
The intelligent prediction method using a fully convolutional neural network improves the accuracy and applicability of coal seam goaf two-zone height predictions, addressing the limitations of existing models by integrating coal mining data for safer and more efficient mining operations.
Patent Information
- Application Number
- CN202411322538.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-23
- Publication Date
- 2024-12-17
- Estimated Expiration
- 2044-09-23
AI Technical Summary
Existing coal seam goaf two-zone height prediction models are not universal and do not fully utilize known data, making it difficult to effectively guide mining operations and prevent underground water hazards.
An intelligent prediction method using a fully convolutional neural network model that utilizes coal mining data to classify areas, construct training sets, and optimize predictions based on measured data, incorporating geological, hydrological, and mechanical parameters.
Enhances the accuracy and universality of two-zone height predictions, facilitating safer and more efficient coal mining by leveraging existing data and reducing prediction complexity.
Smart Images
Figure CN118840216B_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the technical field of coal seam two-zone height prediction, and specifically relates to the design of a two-zone height intelligent prediction method for a coal mine goaf. Background Art
[0002] "Safety, greenness and efficiency" are important development directions and requirements for energy resource development. The goaf fissures formed in the goaf during coal mining have a significant impact on coal mining, ground structures and geological environment. Roof delamination water hazards directly related to the movement and damage of the overburden in the goaf also pose a huge threat to coal mine safety production. The study of the collapse zone and water-conducting fissure zone (hereinafter referred to as the two zones) formed by the movement and damage of the overburden in the goaf is of great significance to the construction of green mines.
[0003] At present, the actual measurement of the height of the two zones is aimed at the areas where the goaf has been formed. The results are not enough to guide the next step of mining work, and it is difficult to prevent and control water underground. It is more urgent for major coal companies to make predictions and prejudgments at the front end to guide the next step of coal mine water prevention and control. The prediction of the development of the two zones is mainly based on empirical formulas and prediction models summarized by some scholars. The prediction models proposed by different scholars based on their own conditions are also different. The existing prediction model of the height of the two zones of coal seam goaf is not universal. Summary of the invention
[0004] The purpose of the present invention is to solve the problem that the existing coal seam goaf two-zone height prediction model does not make full use of known data and is not universal, and proposes an intelligent prediction method for the two-zone height of coal mine goaf.
[0005] The technical solution of the present invention is: a method for intelligently predicting the height of two zones in a coal mine goaf, comprising the following steps:
[0006] S1. Obtain coal mining data in coal mining areas.
[0007] S2. Classify the coal mining areas according to the characteristics of the coal mining strata.
[0008] S3. Based on coal mining data, construct existing data for each coal mining area.
[0009] S4. Construct a training set of the fully convolutional neural network model based on the existing data.
[0010] S5. Input the training set into the fully convolutional neural network model for training to obtain a two-zone height prediction model for the coal mine goaf.
[0011] S6. Input the coal mining design data into the two-zone height prediction model of the coal mine goaf to obtain the two-zone height prediction results of the coal mine goaf.
[0012] S7. Collect the measured data of the two-zone height in the goaf of the coal mine, and optimize the training set through the measured data of the two-zone height.
[0013] S8. Verify the predicted results of the two-zone heights through the measured data of the two-zone heights, and adjust the coal mining area classification results based on the verification results.
[0014] Furthermore, the coal mining data of the S1 coal mining area include geological structure, coal mining stratum characteristics of the main coal mining areas, hydrogeological data, mechanical property data, excavation data, coal resource occurrence characteristic data, mining time data and measured data of the existing two belt heights.
[0015] Furthermore, S3 includes the following sub-steps:
[0016] S31. Obtain model parameters for each coal mining area based on coal mining data. The model parameters include hydrogeological characteristics, mechanical characteristics, coal seam inclination, coal seam burial depth, coal seam mining thickness, overburden structure, vertex coordinates of the completed working face, mining time, existing measured caving zone height and existing measured fracture zone height.
[0017] S32, gridding the model parameters in space to obtain the existing data of the nth sample point, the existing data including the spatial coordinates , Hydrogeological characteristic parameters , mechanical characteristic parameters , coal seam inclination , Coal seam depth , Coal seam mining thickness , Overburden structure , Vertex coordinates of the completed working face , Mining time , the measured height of the caving zone and the measured fracture zone height , .
[0018] Further, S4 is specifically: Convert to a × ×9 three-dimensional data volume as the input of the full convolutional neural network model. Convert to a × ×2 three-dimensional data volume as the label of the full convolutional neural network model, and obtain the training set of the full convolutional neural network model; represents the number of sample points in the east-west direction, Indicates the number of sample points in the north-south direction.
[0019] Further, S6 includes the following sub-steps:
[0020] S61. Grid-number the coal mining design data to obtain the spatial coordinates of the mth sample point. , Hydrogeological characteristic parameters , mechanical characteristic parameters , coal seam inclination , Coal seam depth , Coal seam mining thickness , Overburden structure , Vertex coordinates of the completed working face and mining time , .
[0021] S62, will Convert to a × ×9 three-dimensional data volume, input the two-zone height prediction model of the coal mine goaf, and obtain the predicted value of the caving zone height and fracture zone height prediction .
[0022] S63, according to the predicted value of the caving zone height and fracture zone height prediction , combined with the spatial coordinates of the sample points, the predicted plane diagram of the caving zone height and the predicted plane diagram of the fracture zone height are obtained.
[0023] Further, S7 includes the following sub-steps:
[0024] S71. Collect measured data on the two-zone height of the coal mine goaf through underground transient, micro-resistivity scanning imaging logging or borehole water injection test.
[0025] S72. Optimize the input of the full convolutional neural network model through two-band height measured data.
[0026] Furthermore, S8 is specifically as follows: in response to the error of the two-zone height predicted result based on the two-zone height measured data being greater than a preset threshold, adjusting the coal mining area classification result.
[0027] The beneficial effects of the present invention are:
[0028] (1) The present invention makes full use of the existing coal mining data and has universal applicability for the prediction of the two-zone height of goafs in various types of coal mines, which is conducive to safe and efficient coal mining and green mine construction.
[0029] (2) The present invention pre-classifies the coal mining area according to the characteristics of the coal mining strata, reduces the complexity of the subsequent prediction model, and improves the accuracy of the two-zone height prediction under each specific geological type by sacrificing the generalization ability.
[0030] (3) The present invention adopts a fully convolutional neural network for model feature learning and feature expression, which can fully learn and utilize the existing two-band height measured data. BRIEF DESCRIPTION OF THE DRAWINGS
[0031] Figure 1 The figure shows a flow chart of a method for intelligently predicting the height of two zones in a coal mine goaf provided by an embodiment of the present invention.
[0032] Figure 2 Shown is a schematic diagram of the training of a fully convolutional neural network model provided by an embodiment of the present invention.
[0033] Figure 3 Shown is a prediction schematic diagram of a two-zone height prediction model for a coal mine goaf area provided by an embodiment of the present invention. DETAILED DESCRIPTION
[0034] Now, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings. It should be understood that the embodiments shown and described in the accompanying drawings are only exemplary and are intended to explain the principles and spirit of the present invention, rather than to limit the scope of the present invention.
[0035] The embodiment of the present invention provides a method for intelligently predicting the height of two zones in a coal mine goaf. Figure 1 As shown, the following steps S1 to S8 are included:
[0036] S1. Obtain coal mining data in coal mining areas.
[0037] In the embodiment of the present invention, the coal mining data of the coal mining area includes geological structure, coal mining stratum characteristics of the main coal mining area, hydrogeological data, mechanical property data, mining data, coal resource occurrence characteristic data, mining time data and existing two-belt height measured data.
[0038] S2. Classify the coal mining areas according to the characteristics of the coal mining strata.
[0039] In the embodiment of the present invention, taking the Sichuan coal mining area as an example, according to the coal mining stratum characteristics in the coal mining data, the Sichuan coal mining area is divided into typical coal mining areas such as Panxi area, southern Sichuan area, northern Sichuan area, eastern Sichuan area, and central Sichuan area.
[0040] S3. Based on coal mining data, construct existing data for each coal mining area.
[0041] S3 includes the following sub-steps S31-S32:
[0042] S31. Obtain model parameters for each coal mining area based on coal mining data. The model parameters include hydrogeological characteristics, mechanical characteristics, coal seam inclination, coal seam burial depth, coal seam mining thickness, overburden structure, vertex coordinates of the completed working face, mining time, existing measured caving zone height and existing measured fracture zone height.
[0043] S32, gridding the model parameters in space to obtain the existing data of the nth sample point, the existing data including the spatial coordinates , Hydrogeological characteristic parameters , mechanical characteristic parameters , coal seam inclination , Coal seam depth , Coal seam mining thickness , Overburden structure , Vertex coordinates of the completed working face , Mining time , the measured height of the caving zone and the measured fracture zone height , .
[0044] S4. Construct a training set of the fully convolutional neural network model based on the existing data.
[0045] In the embodiment of the present invention, Convert to a × ×9 three-dimensional data volume as the input of the full convolutional neural network model. Convert to a × ×2 three-dimensional data volume as the label of the full convolutional neural network model, and obtain the training set of the full convolutional neural network model; represents the number of sample points in the east-west direction, Indicates the number of sample points in the north-south direction.
[0046] S5, such as Figure 2 As shown in the figure, the training set is input into the fully convolutional neural network model for training, and a two-zone height prediction model for the coal mine goaf is obtained.
[0047] S6. Input the coal mining design data into the two-zone height prediction model of the coal mine goaf to obtain the two-zone height prediction results of the coal mine goaf.
[0048] S6 includes the following sub-steps S61-S63:
[0049] S61. Grid-number the coal mining design data to obtain the spatial coordinates of the mth sample point. , Hydrogeological characteristic parameters , mechanical characteristic parameters , coal seam inclination , Coal seam depth , Coal seam mining thickness , Overburden structure , Vertex coordinates of the completed working face and mining time , .
[0050] S62, will Convert to a × ×9 three-dimensional data volume, input the two-zone height prediction model of the coal mine goaf, and obtain the predicted value of the caving zone height and fracture zone height prediction ,like Figure 3 shown.
[0051] S63, according to the predicted value of the caving zone height and fracture zone height prediction , combined with the spatial coordinates of the sample points, the predicted plane diagram of the caving zone height and the predicted plane diagram of the fracture zone height are obtained.
[0052] S7. Collect the measured data of the two-zone height in the goaf of the coal mine, and optimize the training set through the measured data of the two-zone height.
[0053] S7 includes the following sub-steps S71-S72:
[0054] S71. Collect measured data on the two-zone height of the coal mine goaf through underground transient, micro-resistivity scanning imaging logging or borehole water injection test.
[0055] S72. Optimize the input of the full convolutional neural network model through two-band height measured data to enrich the training set.
[0056] S8. Verify the predicted results of the two-zone heights through the measured data of the two-zone heights, and adjust the coal mining area classification results based on the verification results.
[0057] In the embodiment of the present invention, in response to the error of the two-zone height measured data to the two-zone height prediction result being greater than a preset threshold, the coal mining area classification result is adjusted.
[0058] Those skilled in the art will appreciate that the embodiments described herein are intended to help readers understand the principles of the present invention, and should be understood that the protection scope of the present invention is not limited to such specific statements and embodiments. Those skilled in the art can make various other specific variations and combinations that do not deviate from the essence of the present invention based on the technical revelations disclosed by the present invention, and these variations and combinations are still within the protection scope of the present invention.
Claims
1. A method for intelligently predicting the height of two zones in a coal mine goaf, characterized in that: The following steps are involved: S1. Obtain coal mining data in coal mining areas; S2. Classify the coal mining areas according to the characteristics of the coal mining strata; S3. Based on the coal mining data, construct the existing data of each coal mining area; S4, construct a training set of a fully convolutional neural network model based on the existing data; S5, inputting the training set into the fully convolutional neural network model for training, and obtaining a two-zone height prediction model for the coal mine goaf; S6. Input the coal mining design data into the two-zone height prediction model of the coal mine goaf to obtain the two-zone height prediction results of the coal mine goaf; S7, collecting two-zone height measured data of the coal mine goaf, and optimizing the training set through the two-zone height measured data; S8. Verify the predicted results of the two-zone heights through the measured data of the two-zone heights, and adjust the coal mining area classification results based on the verification results; The S3 comprises the following sub-steps: S31, obtaining model parameters of each coal mining area according to the coal mining data, wherein the model parameters include hydrogeological characteristics, mechanical characteristics, coal seam inclination, coal seam burial depth, coal seam mining thickness, overburden structure, vertex coordinates of the completed working face, goaf time, existing measured caving zone height and existing measured fracture zone height; S32, spatially gridding the model parameters to obtain existing data of the nth sample point, wherein the existing data includes spatial coordinates , Hydrogeological characteristic parameters , mechanical characteristic parameters , coal seam inclination , Coal seam depth , Coal seam mining thickness , Overburden structure , Vertex coordinates of the completed working face , Mining time , the measured height of the caving zone and the measured fracture zone height , .
2. The method for intelligently predicting the height of two zones in a coal mine goaf according to claim 1 is characterized in that: The coal mining data of the S1 coal mining area include geological structure, coal mining stratum characteristics of the main coal mining areas, hydrogeological data, mechanical property data, mining data, coal resource occurrence characteristic data, mining time data and existing two-belt height measured data.
3. The method for intelligently predicting the height of two zones in a coal mine goaf according to claim 1 is characterized in that: The S4 is specifically: Convert to a × ×9 three-dimensional data volume as the input of the full convolutional neural network model. Convert to a × ×2 three-dimensional data volume as the label of the full convolutional neural network model, and obtain the training set of the full convolutional neural network model; represents the number of sample points in the east-west direction, Indicates the number of sample points in the north-south direction.
4. The method for intelligently predicting the height of two zones in a coal mine goaf according to claim 3 is characterized in that: The S6 comprises the following sub-steps: S61. Grid-number the coal mining design data to obtain the spatial coordinates of the mth sample point. , Hydrogeological characteristic parameters , mechanical characteristic parameters , coal seam inclination , Coal seam depth , Coal seam mining thickness , Overburden structure , Vertex coordinates of the completed working face and mining time , ; S62, will Convert to a × ×9 three-dimensional data volume, input the two-zone height prediction model of the coal mine goaf, and obtain the predicted value of the caving zone height and fracture zone height prediction ; S63, according to the predicted value of the caving zone height and fracture zone height prediction , combined with the spatial coordinates of the sample points, the predicted plane diagram of the caving zone height and the predicted plane diagram of the fracture zone height are obtained.
5. The method for intelligently predicting the height of two zones in a coal mine goaf according to claim 3 is characterized in that: The S7 comprises the following sub-steps: S71. Collect the measured data of the two-zone height of the coal mine goaf through underground transient, micro-resistivity scanning imaging logging or borehole water injection test; S72. Optimize the input of the full convolutional neural network model through two-band height measured data.
6. The method for intelligently predicting the height of two zones in a coal mine goaf according to claim 1 is characterized in that: The S8 is specifically as follows: in response to the error of the two-zone height measured data to the two-zone height prediction result being greater than a preset threshold, adjusting the coal mining area classification result.