Evaluation method for association degree between long observation hole water level and working face mining

Abstract

The invention discloses a long observation hole water level and working face mining association degree evaluation method, which belongs to the technical field of mineral engineering, and comprises the following steps: step 1, collecting water level response data as a basic data set; 2, drawing a long observation hole water level daily variation curve of the basic data set, and constructing a deep learning training total data set; 3, screening the deep learning training total data set to obtain a deep learning training data set; step 4, constructing a feature learning deep neural network model, and using a deep learning training data set to complete training; 5, in the recovery process of the to-be-evaluated working face, the feature learning deep neural network model is used for conducting relevancy classification on the peripheral long observation hole water level daily variation curve; and step 6, performing staged evaluation on the long observation hole water level and the mining association degree of the to-be-evaluated working face. According to the method, the confined aquifer long observation hole with better early warning applicability can be selected by evaluating the correlation degree between the water level fluctuation of the long observation hole and the mining of the to-be-evaluated working face.

Description

Technical Field

[0001] The present invention belongs to the technical field of mining engineering, and particularly relates to a method for evaluating the correlation degree between the water level of a long-term observation hole and the mining movement of a working face. Background Art

[0002] At present, some workable Permian coal seams are hidden under a thick Cenozoic loose layer. The Cenozoic loose pore confined aquifer (mainly the fourth aquifer, hereinafter referred to as the "fourth aquifer") has a large thickness and high water pressure. Under certain mining conditions, there is a significant risk of water inrush disasters in the coal mining working face, seriously threatening the lives of miners and the normal operation of coal mines. Before coal mining operations, establishing a water inrush monitoring and early warning system and a disposal plan for working faces with the risk of water inrush from the "fourth aquifer" is an important guarantee for realizing safe production.

[0003] Engineering practice shows that the water level change of the "fourth aquifer" confined aquifer can reflect the movement characteristics of the overlying strata during mining in the working face. When the development of the water-conducting fissure zone in the working face does not affect the confined aquifer, the linkage relationship between the mining movement in the working face and the water level of the "fourth aquifer" is as follows: the water level of the long-term observation hole shows an obvious single-day fluctuating rise with the mining of the working face. That is, when the production shift is in the mining operation, the roof pressure is relieved and the water level drops. When the maintenance shift stops mining, the "fourth aquifer" obtains recharge and the water level rises. The amplitude of the fluctuation varies due to the characteristics of the aquifer and the mining state. That is, during the safe mining process of the working face, the water level of the "fourth aquifer" above it changes in linkage with the mining operation of the working face. When it suddenly deviates from this linkage change rule, it indicates that a disaster may occur. Based on this change, the staff carry out water inrush early warning work. For example, when the water level of the confined aquifer continues to drop without rising, it indicates that there may be a water inrush phenomenon where a large amount of groundwater suddenly rushes into the mine roadway, and a water inrush early warning needs to be issued, and the staff conduct relevant investigations to prevent the occurrence of water inrush disasters.

[0004] The water level of the "fourth aquifer" is obtained through the hydrogeological long-term observation holes arranged in the mining area. During on-site mining operations, the water level of the long-term observation hole is an important indicator for the risk warning of water inrush from the "fourth aquifer" to support the hydraulic support. Under complex hydrogeological conditions, the response degrees of the water levels of multiple "fourth aquifer" long-term observation holes around the working face to the movement of the overlying strata are not consistent. When the correlation degree between the water level of the "fourth aquifer" long-term observation hole and the mining movement in the working face is strong, the change in its water level has more reference value for water inrush early warning. Conversely, false alarms and missed alarms will occur, affecting the normal production process or losing the early warning function.

[0005] However, there is currently no evaluation method for evaluating the strength of the correlation between the water level of the long-term observation hole and the mining movement in coal mines. As a result, many mines do not evaluate the strength of the correlation between the water level of the long-term observation hole and the mining movement in the working face when constructing a water inrush early warning system, and it is easy to have a situation where the applicability of the long-term observation hole is not strong, resulting in the failure to fully utilize the early warning function of its water level change. Summary of the Invention

[0006] The object of the present invention is to overcome the deficiencies of the above-mentioned prior art and provide a method for evaluating the correlation between the long-term observation hole water level and the working face mining movement.

[0007] To achieve the above object, the present invention adopts the following technical solutions: A method for evaluating the correlation between the long-term observation hole water level and the working face mining movement, comprising the following steps: Step 1, before the mining of the working face to be evaluated, collect the monitoring data of the operation process of the mined working face with similar conditions in the vicinity and the response of the long-term observation hole water level as the basic data set; Step 2, draw the daily change curve of the long-term observation hole water level of the basic data set, and combine the operation process of the mined working face to mark the daily change curve of the long-term observation hole water level of the basic data set as the strong correlation response curve and the weak correlation response curve of the working face mining movement, and construct the total deep learning training data set; Step 3, screen the data of the total deep learning training data set to obtain the deep learning training data set; Step 4, construct a feature learning deep neural network model and complete the training using the deep learning training data set; Step 5, during the mining process of the working face to be evaluated, use the feature learning deep neural network model to classify the correlation degree of the daily change curve of the long-term observation hole water level around; Step 6, according to the correlation degree classification result, conduct a phased evaluation of the correlation between the long-term observation hole water level and the mining movement of the working face to be evaluated.

[0008] Preferably, in the step 1, the mined working face with similar conditions in the vicinity is the working face that is in the same coal mine or the same mining area as the working face to be evaluated, has the same mined coal seam, the same confined aquifer, and has been safely mined; the operation process refers to the production shift and the maintenance shift time every day during the mining of the mined working face, and the operation process of the working face to be evaluated is the same as that of the mined working face.

[0009] Preferably, in the step 2, the method of marking the daily change curve of the long-term observation hole water level of the basic data set as the strong correlation response curve and the weak correlation response curve of the working face mining movement is: during the process of coal mining in the single-day production shift - stop mining in the maintenance shift, when the daily change curve of the long-term observation hole water level in the confined aquifer conforms to the law of single-day fluctuating rise, mark the daily change curve as the strong correlation response curve of the working face mining movement, otherwise mark it as the weak correlation response curve of the working face mining movement; The law of single-day fluctuating rise is: the water level of the underground long-term observation hole drops during coal mining in the production shift and rises during stop mining in the maintenance shift.

[0010] Preferably, in the step 3, in the deep learning training data set, the proportion of the data of the strong correlation response curve of the working face mining movement is 70%, and the proportion of the data of the weak correlation response curve of the working face mining movement is 30%.

[0011] Optimized. In step 4, the feature learning deep neural network model structure includes an input layer, a convolutional layer A, a pooling layer A, a convolutional layer B, a pooling layer B, and a fully connected layer. The input end of the input layer is used to input the long-term observation hole daily variation curve graph. The output end of the input layer is communicatively connected to the input end of the convolutional layer A. The output end of the convolutional layer A is communicatively connected to the input end of the pooling layer A. The output end of the pooling layer A is communicatively connected to the input end of the convolutional layer B. The output end of the convolutional layer B is communicatively connected to the input end of the pooling layer B. The output end of the pooling layer B is communicatively connected to the input end of the fully connected layer. The fully connected layer is used to output two labels, namely strongly correlated and weakly correlated.

[0012] Preferably, the long-term observation hole daily variation curve graph input in the input layer is 224×224 pixels. Sixteen convolutional kernels with a size of 5×5 and a stride of 1 are arranged in the convolutional layer A. The pooling window size in the pooling layer A is 2×2 and the stride is 2. Thirty-two convolutional kernels with a size of 3×3 and a stride of 1 are arranged in the convolutional layer B. The pooling window size in the pooling layer B is 1×1 and the stride is 1.

[0013] Preferably, in step 4, when using the deep learning training dataset to complete the training, the deep learning training dataset is split into a training set and a validation set according to a ratio. The maximum number of training times is 500 times. The model training achievement conditions: the fitting error of the training set is less than 5%, the fitting error of the validation set is less than 5%, and 20% of the data in the deep learning training dataset is randomly selected twice as the test set, and the fitting error of the test set is less than 5%.

[0014] Preferably, in step 5, after the mining of the working face to be evaluated starts, the daily variation curve of the water level of the surrounding long-term observation holes is classified for correlation using the feature learning deep neural network model in step 4, and is marked as strongly correlated with the mining movement of the working face or weakly correlated with the mining movement of the working face according to the classification result.

[0015] Preferably, step 6 includes the following sub-steps: Step 61, determine the mining distance value of a stage during the phased evaluation of the working face to be evaluated according to the characteristics of the overlying strata change of the coal seam. Step 62, after the mining of the jth (j≥1) stage is completed, conduct a phased evaluation of the correlation between the water level of the long-term observation hole and the mining movement during the mining of the jth stage of the working face to be evaluated. The method is as follows: For each long-term observation borehole, count the number of days during the mining process in the j-th stage when its water level fluctuation has a strong correlation response with the mining of the working face and the number of days when its water level fluctuation has a weak correlation response with the mining of the working face ; For each long-term observation borehole, count the proportion of the number of days when its water level fluctuation has a strong correlation response with the mining of the working face in the number of days of normal mining of the working face in the j-th stage ; ×100% (1) According to the proportion of the number of days when the water level fluctuation has a strong correlation response with the mining of the working face in the number of days of mining in the j-th stage of the working face Classify the strength of the correlation between the water level of the long-term observation borehole and the mining in the j-th stage of the working face to be evaluated

[0016] Preferably, in step 62, for long-term observation boreholes with a proportion ≥70%, their correlation with the mining of the working face to be evaluated is strong, defined as level-I correlation boreholes; for long-term observation boreholes with a proportion of 40≤ <70%, their correlation with the mining of the working face to be evaluated is medium, defined as level-II correlation boreholes; for long-term observation boreholes with a proportion <40%, their correlation with the mining of the working face to be evaluated is weak, defined as level-III correlation boreholes

[0017] The beneficial effects of the present invention are as follows Based on the water level monitoring data of the long-term observation boreholes in the confined aquifer during the mining process of the working face, by evaluating the correlation between the water level fluctuation of the long-term observation boreholes and the mining of the working face to be evaluated, the long-term observation boreholes in the confined aquifer with better early warning applicability can be selected, and during the subsequent mining process of the working face, the water inrush disaster can be more accurately predicted BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The accompanying drawings forming a part of this application are used to provide a further understanding of this application. The illustrative embodiments and descriptions thereof of this application are used to explain this application and do not constitute an improper limitation to this application

[0019] Figure 1 is a schematic flow chart of the method for evaluating the correlation between the water level of the long-term observation borehole and the mining of the working face of the present invention Figure 2 is the strong correlation response curve between the water level of the long-term observation borehole in the confined aquifer and the mining of the working face of the present invention Figure 3 is the weak correlation response curve between the water level of the long-term observation borehole in the confined aquifer and the mining of the working face of the present invention Figure 4 is the stage evaluation result of the correlation between the water level of the long-term observation borehole and the mining of the working face in the embodiment DETAILED DESCRIPTION OF THE INVENTION

[0020] It should be noted that the following detailed description is illustrative and is intended to provide further explanation of the present application. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by those of ordinary skill in the technical field to which the present application belongs.

[0021] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present application. As used herein, unless the context clearly indicates otherwise, the singular forms are also intended to include the plural forms. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0022] In the present invention, terms such as "upper", "lower", "bottom", "top", etc. indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings. They are only relationship terms determined for the convenience of describing the structural relationship of each component or element of the present invention and do not specifically refer to any component or element in the present invention and should not be construed as a limitation on the present invention.

[0023] In the present invention, terms such as "connected" and "coupled" should be understood in a broad sense and may mean a fixed connection, an integral connection or a detachable connection; they may be directly connected or indirectly connected through an intermediate medium. For those related scientific research or technical personnel in the field, the specific meanings of the above terms in the present invention can be determined according to specific circumstances and should not be construed as a limitation on the present invention.

[0024] The present invention will be further described below in conjunction with the drawings and embodiments.

[0025] As Figure 1 shown, a method for evaluating the correlation between the long - term observation hole water level and the mining movement of the working face includes the following steps: Step 1, before the mining of the working face to be evaluated, collect the operation process of the mined working face with similar conditions around and the monitoring data of the long - term observation hole water level response as the basic data set; Step 2, draw the daily change curve of the long - term observation hole water level of the basic data set. Combining with the operation process of the mined working face, mark the daily change curve of the long - term observation hole water level of the basic data set into two categories: the strong - correlation response curve and the weak - correlation response curve of the mining movement of the working face, and construct the total deep - learning training data set; Step 3, perform data screening on the total deep - learning training data set to obtain the deep - learning training data set; Step 4, construct a feature - learning deep neural network model and complete the training using the deep - learning training data set; Step 5: During the mining process of the working face to be evaluated, use the feature learning deep neural network model to classify the correlation degree of the daily variation curves of the water levels of the surrounding long-term observation holes; Step 6: According to the correlation classification results, conduct phased evaluations on the correlation between the water levels of the long-term observation holes and the mining activities of the working face to be evaluated.

[0026] Preferably, in step 1, the mined working faces with similar surrounding conditions are the working faces that are in the same coal mine or the same mining area as the working face to be evaluated, have the same mined coal seam, the same confined aquifer, and have been safely mined; the operation process refers to the production shift and maintenance shift time during the mining of the mined working face, and the operation process of the working face to be evaluated is the same as that of the mined working face.

[0027] Preferably, in step 2, the method of marking the daily variation curves of the water levels of the long-term observation holes in the basic data set as strong-correlation response curves and weak-correlation response curves of working face mining is as follows: During the process of coal mining in the production shift and coal mining suspension in the maintenance shift, when the daily variation curve of the water level of the confined aquifer long-term observation hole conforms to the law of single-day fluctuating rise, mark this daily variation curve as a strong-correlation response curve of working face mining, otherwise mark it as a weak-correlation response curve of working face mining; The law of single-day fluctuating rise is that the water level of the underground long-term observation hole drops during coal mining in the production shift and rises during coal mining suspension in the maintenance shift.

[0028] Among them, the water level of the confined aquifer long-term observation hole and the strong-correlation response curve of working face mining are as Figure 2 shown, and the water level of the confined aquifer long-term observation hole and the weak-correlation response curve of working face mining Figure 3 are shown.

[0029] Preferably, in step 3, the deep learning training data set is screened to make it more conducive to feature training. In the deep learning training data set, the proportion of the data of the strong-correlation response curves of working face mining is 70%, and the proportion of the data of the weak-correlation response curves of working face mining is 30%.

[0030] Optimally, in step 4, the structure of the feature learning deep neural network model includes an input layer, a convolutional layer A, a pooling layer A, a convolutional layer B, a pooling layer B, and a fully connected layer; The input end of the input layer is used to input the daily variation curve graph of the long-term observation hole. The output end of the input layer is communicatively connected to the input end of the convolutional layer A. The output end of the convolutional layer A is communicatively connected to the input end of the pooling layer A. The output end of the pooling layer A is communicatively connected to the input end of the convolutional layer B. The output end of the convolutional layer B is communicatively connected to the input end of the pooling layer B. The output end of the pooling layer B is communicatively connected to the input end of the fully connected layer; The fully connected layer is used to output two labels, namely strong correlation and weak correlation.

[0031] Preferably, the long-term observation hole daily change curve graph input in the input layer is 224×224 pixels; 16 convolution kernels with a size of 5×5 and a stride of 1 are arranged in the convolution layer A; The pooling window size in the pooling layer A is 2×2 and the stride is 2, that is, it moves 2 pixels each time, thereby reducing the dimension of the feature map; 32 convolution kernels with a size of 3×3 and a stride of 1 are arranged in the convolution layer B; The pooling window size in the pooling layer B is 1×1 and the stride is 1.

[0032] Preferably, in step 4, when using the deep learning training data set to complete the special time, the deep learning training data set is divided into a training set and a validation set according to a ratio; among them, the ratio of the number of data in the training set to the number of data in the validation set is 4:1.

[0033] The maximum number of training times is 500 times; The conditions for model training to be achieved: the fitting error of the training set is less than 5%, the fitting error of the validation set is less than 5%, and 20% of the data in the deep learning training data set is randomly selected as the test set for the second time, and the fitting error of the test set is less than 5%.

[0034] Regarding the fitting error, the long-term observation hole daily change curve graph in the training set or the validation set or the test set is input into the feature learning deep neural network model, and the fitting label result is output. The number N of the fitting label results in the corresponding data set that is inconsistent with the situation marked by the long-term observation hole daily change curve graph according to step 2 is counted. Then the fitting error is the proportion of the number N in the number of data in the corresponding data set. For example, each long-term observation hole daily change curve graph in the training set is input into the feature learning deep neural network model, and the fitting label result is output. The number N1 of the fitting label results in the training set that is inconsistent with the situation marked by the long-term observation hole daily change curve graph in the training set according to step 2 is counted. Then the fitting error of the training set is the proportion of the number N1 in the number of data in the training set.

[0035] The inconsistency between the fitting label result and the situation marked by the long-term observation hole daily change curve graph according to step 2 means that: the fitting label result is strongly correlated, and the long-term observation hole daily change curve graph is marked as a weak correlation response curve of the working face mining according to step 2; or, the fitting label result is weakly correlated, and the long-term observation hole daily change curve graph is marked as a strong correlation response curve of the working face mining according to step 2.

[0036] Preferably, in step 5, after the mining of the working face to be evaluated starts, the daily variation curve of the water level of the surrounding long-term observation holes is classified by the feature learning deep neural network model in step 4 every day, and is marked as strongly related to the mining of the working face or weakly related to the mining of the working face according to the classification results.

[0037] Preferably, step 6 includes the following sub-steps: Step 61, determine the mining distance value of a stage during the phased evaluation of the working face to be evaluated according to the characteristics of the overlying strata change of the coal seam, where the mining distance value can be 100 meters; Step 62, after the mining of the jth (j≥1) stage is completed, conduct a phased evaluation of the correlation between the water level of the long-term observation hole and the mining during the mining of the jth stage of the working face to be evaluated. The method is as follows: For each long-term observation hole, count the number of days during the mining of the jth stage when its water level fluctuation is strongly related to the mining of the working face and the number of days when its water level fluctuation is weakly related to the mining of the working face ; For each long-term observation hole, count the proportion of the number of days when its water level fluctuation is strongly related to the mining of the working face in the normal mining days of the working face during the jth stage ; ×100% (1) According to the proportion of the number of days when the water level fluctuation is strongly related to the mining of the working face in the mining days of the jth stage of the working face divide the strength of the correlation between the water level of the long-term observation hole and the mining during the mining of the jth stage of the working face to be evaluated.

[0038] Preferably, in step 62, for long-term observation holes with a proportion ≥70%, their mining is strongly correlated with the working face to be evaluated, and they are defined as level-I correlation holes; for long-term observation holes with a proportion of 40≤ <70%, their mining is moderately correlated with the working face to be evaluated, and they are defined as level-II correlation holes; for long-term observation holes with a proportion <40%, their mining is weakly correlated with the working face to be evaluated, and they are defined as level-III correlation holes.

[0039] Example: Taking the 01 working face in a coal mine mined under the "Fourth Aquifer" confined aquifer as an example, there are four long-term "Fourth Aquifer" water level observation holes, namely Observation 1, Observation 2, Observation 3, and Observation 4, around the 01 working face. The working face operation cycle during the mining of the 01 working face is as follows: The maintenance shift is from 8:00 to 15:00 every day, and the production shift is from 15:00 to 8:00 the next day. The designed mining length is 900 meters.

[0040] Before the mining of the 01 working face, the daily change curves of the water levels of four "Fourth Aquifer" long-term observation holes during the mining of the mined working faces with similar conditions in this area were collected. 560 response curves with strong correlation and 240 response curves with weak correlation between the water levels of the "Fourth Aquifer" long-term observation holes and the mining intensity of the mined working faces were screened to construct a deep learning training data set. Based on this deep learning training data set, the training of the deep neural network model for feature learning of the present invention was completed.

[0041] During the mining of the 01 working face, the water levels of four "Fourth Aquifer" long-term observation holes, namely Observation Well 1, Observation Well 2, Observation Well 3, and Observation Well 4, around the working face were monitored and the daily change curves of the water levels were plotted. Based on the deep neural network model for feature learning, the classification of strong correlation and weak correlation was carried out on the correlation degree between the water level fluctuations of the "Fourth Aquifer" long-term observation holes and the mining of the working face.

[0042] During the mining process of the 01 working face, with 100 meters as a stage, the correlation degree between the water levels of the long-term observation holes and the mining of the working face was evaluated stage by stage. The evaluation results are shown in Figure 4 . Among them, in the first 100-meter stage of the mining of the 01 working face, the strong correlation ratios of Observation Well 1, Observation Well 2, Observation Well 3, and Observation Well 4 were 68%, 10%, 52%, and 55% respectively. According to the correlation degree evaluation method of the present invention, Observation Well 1, Observation Well 3, and Observation Well 4 are Class II correlation degree holes, and their water level fluctuations can play a good role in water inrush early warning; Observation Well 2 is a Class III correlation degree hole, and its role in water inrush early warning for the 01 working face is poor. With the continuous mining of the 01 working face, the correlation degree between the water levels of Observation Well 1, Observation Well 3, and Observation Well 4 and the mining of the 01 working face shows a relatively coherent dynamic adjustment. As shown in Table 1, the correlation degree between Observation Well 2 and the mining of the 01 working face is always small. At the end of the mining stage, the overlying strata tend to be stable, and the daily change curves of all long-term observation holes no longer reflect the mining characteristics.

[0043] Table 1 Statistical table of the phased evaluation results of the correlation degree

[0044] In the actual water inrush early warning plan for the 01 working face operation, the single-day rapid drop value of the water level of the "Fourth Aquifer" long-term observation hole is used as an important early warning parameter. In the current mining stage, referring to the evaluation classification of the mining correlation degree of Observation Well 1, Observation Well 2, Observation Well 3, and Observation Well 4 in the previous stage, different early warning weights are defined, which greatly reduces the false alarm rate and realizes the safe and efficient mining of the 01 working face.

[0045] Although the specific implementation manners of the present invention have been described above in conjunction with the accompanying drawings, they are not limitations on the present invention. Those skilled in the art should understand that various modifications or deformations that can be made without creative efforts on the basis of the technical solutions of the present invention are still within the protection scope of the present invention.

Claims

1. An evaluation method for the correlation between the water level in the long-term observation hole and the mining movement of the working face, characterized in that, It includes the following steps: Step 1: Before the mining of the working face to be evaluated, collect the monitoring data of the operation process and the water level response of the long observation holes of the mined working faces with similar conditions around as the basic data set; Step 2: Draw the daily variation curve of the water level of the long observation holes in the basic data set. Combine the operation process of the mined working face, and mark the daily variation curve of the water level of the long observation holes in the basic data set as the strong correlation response curve and the weak correlation response curve of the working face mining movement, and construct the total deep learning training data set; Step 3: Screen the data in the total deep learning training data set to obtain the deep learning training data set; Step 4: Construct a feature learning deep neural network model and complete the training using the deep learning training data set; Step 5: During the mining process of the working face to be evaluated, use the feature learning deep neural network model to classify the correlation degree of the daily variation curve of the water level of the surrounding long observation holes; Step 6: According to the classification results of the correlation degree, conduct a phased evaluation on the correlation degree between the water level of the long observation hole and the mining movement of the working face to be evaluated.

2. The evaluation method for the correlation degree between the long observation hole water level and the mining movement of the working face according to claim 1, characterized in that, In the above Step 1, the mined working faces with similar conditions around are the working faces that are in the same coal mine or the same mining area as the working face to be evaluated, with the same mined coal seam, the same confined aquifer, and have been safely mined; the operation process refers to the production shift and maintenance shift time every day during the mining of the mined working face, and the operation process of the working face to be evaluated is the same as that of the mined working face.

3. The evaluation method for the correlation degree between the long observation hole water level and the working face mining movement according to claim 1, characterized in that In the above Step 2, the method of marking the daily variation curve of the water level of the long observation holes in the basic data set as the strong correlation response curve and the weak correlation response curve of the working face mining movement is: during the process of coal mining in the single-day production shift and coal mining stop in the maintenance shift, when the daily variation curve of the water level of the long observation hole in the confined aquifer conforms to the law of single-day fluctuating rise, mark this daily variation curve as the strong correlation response curve of the working face mining movement, otherwise mark it as the weak correlation response curve of the working face mining movement; The law of single-day fluctuating rise is: the water level of the underground long observation hole drops during coal mining in the production shift and rises during coal mining stop in the maintenance shift.

4. The evaluation method for the correlation degree between the water level of the long observation hole and the mining movement of the working face according to claim 3, characterized in that, In the above Step 3, in the deep learning training data set, the proportion of the data of the strong correlation response curve of the working face mining movement is 70%, and the proportion of the data of the weak correlation response curve of the working face mining movement is 30%.

5. The evaluation method for the correlation degree between the long observation hole water level and the working face mining movement according to claim 3, characterized in that, In the above Step 4, the structure of the feature learning deep neural network model includes an input layer, a convolutional layer A, a pooling layer A, a convolutional layer B, a pooling layer B, and a fully connected layer; The input end of the input layer is used to input the daily variation curve graph of the long observation hole. The output end of the input layer is communicatively connected to the input end of the convolutional layer A. The output end of the convolutional layer A is communicatively connected to the input end of the pooling layer A. The output end of the pooling layer A is communicatively connected to the input end of the convolutional layer B. The output end of the convolutional layer B is communicatively connected to the input end of the pooling layer B. The output end of the pooling layer B is communicatively connected to the input end of the fully connected layer; The fully connected layer is used to output two labels, namely strong correlation and weak correlation.

6. The evaluation method for the correlation degree between the long observation hole water level and the working face mining movement according to claim 5, wherein, The daily variation curve graph of the long observation hole input in the input layer is 224×224 pixels; In the convolutional layer A, 16 convolutional kernels with a size of 5×5 are arranged, and the stride is 1; In the pooling layer A, the size of the pooling window is 2×2, and the stride is 2; 32 convolutional kernels of size 3×3 are arranged in the convolutional layer B, with a stride of 1; The pooling window size in the pooling layer B is 1×1, and the stride is 1.

7. The evaluation method for the correlation degree between the long observation hole water level and the working face mining movement according to claim 6, characterized in that, In step 4, when using the deep learning training dataset to complete the training, the deep learning training dataset is proportionally divided into a training set and a validation set; The maximum number of training times is 500 times; The conditions for model training to be achieved are: the fitting error of the training set is less than 5%, the fitting error of the validation set is less than 5%, 20% of the data in the deep learning training dataset is randomly selected twice as the test set, and the fitting error of the test set is less than 5%.

8. The evaluation method for the correlation degree between the long observation hole water level and the working face mining movement according to claim 3, characterized in that, In step 5, after the mining of the working face to be evaluated starts, the daily variation curve of the water level of the surrounding long-term observation holes is classified by correlation using the feature learning deep neural network model in step 4 every day, and is marked as strongly related to the mining of the working face or weakly related to the mining of the working face according to the classification results.

9. The evaluation method for the correlation degree between the water level in the long observation hole and the mining movement of the working face according to claim 8, characterized in that, Step 6 includes the following sub-steps: Step 61, determine the mining distance value of a stage during the staged evaluation of the working face to be evaluated according to the characteristics of the overlying strata change of the coal seam; Step 62, after the mining of the jth (j≥1) stage is completed, conduct a staged evaluation of the correlation between the water level of the long-term observation hole and the mining during the mining of the jth stage of the working face to be evaluated. The method is as follows: For each long-term observation borehole, count the number of days during the mining process in the j-th stage when its water level fluctuation has a strong correlation response with the mining of the working face and the number of days when its water level fluctuation has a weak correlation response with the mining of the working face ; For each long observation hole, count the proportion of the number of days with water level fluctuations strongly correlated with the mining activities of the working face in the number of days of normal coal mining of the working face in the j-th stage ; ×100% （1） Percentage of the number of days with a strong correlation between water level fluctuations and mining activities in the working face in the number of days of the j-th stage of mining in the working face Divide the water level of the long-term observation borehole into strong and weak degrees of mining correlation during the mining process of the j-th stage of the working face to be evaluated.

10. The evaluation method for the correlation between the long-view hole water level and the mining movement of the working face according to claim 9, characterized in that, In the said step 62, the proportion ≥70% of the long observation holes, which have a strong correlation with the mining of the working face to be evaluated, are defined as level-I correlation holes; the proportion is 40 ≤ <70% of the long observation holes, which have a medium correlation with the mining of the working face to be evaluated, are defined as level-II correlation holes; the proportion <40% of the long observation holes, which have a weak correlation with the mining of the working face to be evaluated, are defined as level-III correlation holes.

Images