A method for joint detection of remaining coal thickness in roof by seismic and transient electromagnetic methods

Abstract

The present invention discloses a method for jointly detecting the remaining coal thickness of the roof by seismic and transient electromagnetic methods. First, seismic data and transient electromagnetic data are respectively collected. After the collected seismic data is converted, the formation wave velocity is obtained. Then, the formation wave velocity is fuzzified, and the membership function of the formation wave velocity is optimized. Finally, an optimized seismic data fuzzy set is obtained. The collected transient electromagnetic data is converted into the transient electromagnetic full-term apparent resistivity, and after fuzzification, a transient electromagnetic data fuzzy set is obtained. Then, the fuzzy sets formed respectively are fused by automatically adjusting the weight and membership degree parameters according to different geological conditions. Finally, the geological conditions are obtained based on the fused data, and during the data fusion process, through optimization processing, the fused data can effectively reduce noise interference and improve the signal-to-noise ratio, and finally obtain the optimal fused data, so as to more accurately detect the underground structure information and effectively ensure the accuracy of the detection result.

Description

Technical Field

[0001] The present invention relates to a method for detecting the remaining coal thickness of the roadway roof, specifically a method for jointly detecting the remaining coal thickness of the roof by seismic and transient electromagnetic methods, belonging to the technical field of coal thickness detection in coal mines. Background Art

[0002] Both gas and ground stress are one of the main disaster-causing geological factors in coal mines, and the coal seam thickness has a strong control effect on them, affecting the driving of coal mine roadways and the production safety of working faces. In addition, due to different coal seam occurrence conditions, there are significant differences in the thickness of the top coal. When caving mining is carried out according to a unified coal caving procedure, it often leads to uneven remaining coal thickness of the top coal, affecting coal production. Therefore, finding out the coal seam thickness is an urgent problem to ensure the safety and output of working face mining.

[0003] Traditional means for coal thickness exploration mainly include drilling, roadway exploration and geophysical exploration. Drilling has high precision but limited control range; roadway exploration has too high cost; geophysical exploration, as a non-contact detection method, has the advantages of wide detection range, fast speed and low cost. In the existing technical means, underground coal thickness detection mainly uses single geophysical exploration means such as trough wave seismic and mine seismic waves. Due to the defects of these single methods, the overall detection effect is poor. Therefore, there are also some methods that can improve the detection effect by jointly detecting seismic and transient electromagnetic and verifying each other; however, in the existing joint method of the two, mainly the results are obtained through independent detection respectively, and then the same detection parts are verified with each other to directly determine the geological conditions of this part. For different parts, generally one detection result is selected as the geological conditions of different parts, and both detection methods may have inaccurate detection situations. Only taking one of the detection results for different parts will reduce the accuracy of the detection results. Since the current method cannot fuse the two kinds of data well, a better detection result cannot be obtained through the fused data.

[0004] Therefore, how to provide a method that can automatically adjust the weight and membership degree parameters according to different geological conditions for the data obtained by seismic detection and transient electromagnetic detection respectively, and finally obtain the geological conditions according to the fused data, effectively ensuring the accuracy of the detection results, is one of the research directions in this industry. Summary of the Invention

[0005] Aiming at the problems existing in the above-mentioned prior art, the present invention provides a method for jointly detecting the remaining coal thickness of the roof by seismic and transient electromagnetic methods, which can automatically adjust the weight and membership degree parameters according to different geological conditions for the data obtained by seismic detection and transient electromagnetic detection respectively, and finally obtain the geological conditions according to the fused data, effectively ensuring the accuracy of the detection results.

[0006] To achieve the above object, the technical solution adopted by the present invention is: a method for jointly detecting the remaining coal thickness of the roof by seismic and transient electromagnetic methods, and the specific steps are as follows:

[0007] Step 1. Assemble the seismic and transient electromagnetic joint detection acquisition device: The detection acquisition device includes a T-shaped bracket, a seismic transient acquisition station, an electromagnetic exciter, a plurality of three-component piezoelectric sensors, a plurality of three-component electromagnetic coils, and a control and analysis terminal. The seismic transient acquisition station is connected to the control and analysis terminal; when performing seismic detection, a plurality of three-component piezoelectric sensors and electromagnetic exciters are detachably mounted on the upper part of the T-shaped bracket. The plurality of three-component piezoelectric sensors and electromagnetic exciters are arranged at equal intervals in a straight line, and the electromagnetic exciter is located in the middle of the plurality of three-component piezoelectric sensors; each three-component piezoelectric sensor and electromagnetic exciter are connected to the seismic transient acquisition station; when performing transient electromagnetic detection, a plurality of three-component electromagnetic coils are detachably mounted on the upper part of the T-shaped bracket, and the installation positions of the plurality of three-component electromagnetic coils correspond to the installation positions of each three-component piezoelectric sensor one by one; and each three-component electromagnetic coil is connected to the seismic transient acquisition station;

[0008] Step 2. Establish an observation system for detecting the remaining coal thickness of the roof: Taking the roadway running direction as the X direction, the roadway horizontal plane direction as the Y direction, and the detection depth direction as the Z direction, establish an observation system for detecting the remaining coal thickness of the roof;

[0009] Step 3. Deploy the seismic and transient electromagnetic joint detection acquisition device: First, determine the position where the coal thickness needs to be detected, then install a plurality of three-component piezoelectric sensors and electromagnetic exciters, and then lift the T-shaped bracket until each three-component piezoelectric sensor and electromagnetic exciter are tightly coupled with the coal wall at the top of the roadway, and the arrangement direction of the plurality of three-component piezoelectric sensors and electromagnetic exciters is along the roadway running direction, and the detection direction of the plurality of three-component piezoelectric sensors is towards the top coal seam. At this time, the deployment work is completed;

[0010] Step 4: Collect seismic data and transient electromagnetic data: When conducting seismic exploration, the control and analysis terminal controls the electromagnetic vibrator to start through the seismic transient acquisition station and generate seismic waves at its location. Each three-component piezoelectric sensor receives the seismic wave data reflected from the coal seam and transmits the seismic wave data to the control and analysis terminal through the seismic transient acquisition station. After the seismic data collection is completed, lower the T-shaped support, disassemble each three-component piezoelectric sensor and the electromagnetic vibrator from the T-shaped support, then install multiple three-component electromagnetic coils at the installation positions of the respective three-component piezoelectric sensors one by one. After completion, raise the T-shaped support again until each three-component electromagnetic coil is tightly coupled with the coal wall at the top of the roadway. At this time, the layout work for transient electromagnetic exploration is completed. When conducting transient electromagnetic exploration, the control and analysis terminal controls one of the three-component electromagnetic coils to generate transient electromagnetic signals through the seismic transient acquisition station, and then all three-component electromagnetic coils respectively receive the transient electromagnetic data reflected from the coal seam and transmit the transient electromagnetic data to the control and analysis terminal through the seismic transient acquisition station, thus completing the process of collecting seismic data and transient electromagnetic data.

[0011] Step 5: Preliminarily process the seismic data: Taking the position of the electromagnetic vibrator as the origin, establish a plane XZ coordinate system. The X and Z directions of the XZ coordinate system are consistent with the observation system. Divide the coordinate system with a square grid with a side length of 0.1 m, and place the obtained seismic data into the above XZ coordinate system. According to the time t of the first seismic signal received by each grid where the three-component piezoelectric sensors are located, combined with its grid coordinate position, calculate the seismic apparent velocity V received by each three-component piezoelectric sensor. s ; Then, conduct polarization analysis based on the seismic data of different components, and convert each seismic apparent velocity V s into the formation wave velocity V in the main polarization direction d . Finally, perform fuzzification processing on the formation wave velocity V d , and optimize the membership function of the formation wave velocity to finally obtain the optimized seismic data fuzzy set A1 = f1(V d ), where f1 is the membership function of the formation wave velocity.

[0012] Step 6: Preliminarily process the transient electromagnetic data: Convert the voltage signals collected by each component in the collected transient electromagnetic data into the full-period apparent resistivity ρ s according to the conventional process, and perform fuzzification processing on the transient electromagnetic full-period apparent resistivity data ρ s to obtain the transient electromagnetic data fuzzy set A2 = f2(ρ s ), where f2 is the membership function of the transient electromagnetic full-period apparent resistivity ρ s ; Since the conventional coal thickness is basically in the early stage of electromagnetic wave attenuation for transient electromagnetic, and the influence of small structures such as the formation structure during this period is relatively small, the conventional triangular membership function can be used to achieve the fuzzification processing of the transient electromagnetic data.

[0013] Step 7: Fuse seismic data and transient electromagnetic data: Construct a fuzzy logic rule base containing 2 rules; Input the data after fuzzification in Steps 5 and 6 into the rule base for inference to obtain a fuzzy output set B n = min{A 1i , A 2j}, where n represents the membership degree of the nth output; Convert the fuzzy output set B n into the actual weight value w n , w n = ∑(i,j)B n * w(i,j) / ∑(i,j)B n . Since small structures are contained in the coal seam, resulting in local anomalies in formation velocity, and the lithology and water content of the overlying strata of the coal seam will interfere with the detection data of transient electromagnetic, it will be difficult to overcome the bottleneck of the method itself and improve the accuracy when using the two methods alone. To meet the needs of detecting coal thickness under various geological conditions as much as possible, adaptive improvement is required for the data fusion of the two, so as to automatically adjust the weight and membership degree parameters to match different geological situations. The specific optimization objective function T is as follows:

[0014] T = w i * RMSE(V R , V d , f1) + w j * RMSE(V R , ρ s , f2)

[0015] where V R is the fused data, V d is the formation wave velocity data, f1 is the membership function of the formation wave velocity, w i is the weight corresponding to the formation wave velocity, ρ s is the transient electromagnetic full - period apparent resistivity data, f2 is the membership function of the transient electromagnetic full - period apparent resistivity ρ s , w j is the weight corresponding to the transient electromagnetic full - period apparent resistivity, and RMSE represents the root - mean - square error, specifically:

[0016] RMSE(V R , V d , f1) = (Σ(V Rk - V dk )^2 / k)^0.5

[0017] where V Rk is the kth data point after fusion, V dk is the kth formation wave velocity data, and k is the total number of data points. Based on this optimization objective function T, use an optimization algorithm to find the optimal weight wi , w j and the membership function parameters A, B, C, D up , D down , so that the objective function reaches the minimum value. Finally, weighted averaging is performed according to the optimized weight values corresponding to each data source to obtain the fused data; the data is fused by adaptively adjusting the weights and membership parameters, thereby ensuring the reliability and adaptability of the fusion algorithm.

[0018] Step Eight: Determine the remaining coal thickness: Use the fused data as formation parameters for conventional data inversion, and obtain the remaining coal thickness of the roof according to the inversion result.

[0019] Furthermore, the distance between the multiple three-component piezoelectric sensors and the electromagnetic exciters is 0.2 m.

[0020] Furthermore, the optimization in Step Five is specifically as follows: Due to the complexity and heterogeneity of the formation, the formation wave velocity V d usually does not show a simple linear relationship. Therefore, it is necessary to optimize the conventional membership function to handle this non-linear relationship and more accurately predict the distribution of the formation wave velocity V d . The hyperbolic tangent function is selected to optimize the membership function to describe the formation wave velocity V d data with non-linear growth and saturation characteristics. The specific expression is:

[0021]

[0022] where A and B are the minimum and maximum values of the membership function respectively, C is the parameter of the function change rate; D is the translation amount of the function, and D up and D down represent the translation amounts of the rising section and the falling section respectively.

[0023] Furthermore, the specific calculation process of the seismic apparent velocity in Step Five is as follows:

[0024] Assume that the coordinates of two adjacent piezoelectric sensors are (x1, y1, z1) and (x2, y2, z2) respectively, and their corresponding seismic apparent velocity is x, y, z are the coordinates of the grid where the piezoelectric sensor is located, and t is the time of the first seismic signal recorded by each piezoelectric sensor.

[0025] Furthermore, the rules in the fuzzy logic rule base in Step Seven need to be set in combination with the on-site formation conditions.

[0026] Furthermore, the fuzzy output set B nIt represents the membership degree of the inference result, which is composed of the fuzzy set A1 of seismic data, the fuzzy set A2 of transient electromagnetic data, and their respective membership degree values, and is obtained according to the fuzzy intersection operation.

[0027] Compared with the prior art, since seismic and transient electromagnetic methods are two different geophysical exploration methods, they respectively utilize the propagation characteristics of seismic waves and electromagnetic waves in underground media to detect underground structure information; combining these two methods can obtain underground structure information more accurately; there are differences in the detection objects, detection depths, and physical mechanisms of the two geophysical exploration methods, which can complement each other's deficiencies; and the positioning accuracy is higher than that of a single geophysical exploration method; at the same time, the present invention can overcome the problem that the two cannot be well integrated in the prior art. By first performing fuzzy processing on the data obtained from seismic detection and transient electromagnetic detection respectively, and then the formed fuzzy data sets automatically adjust the weight and membership degree parameters according to different geological conditions for fusion, and finally obtain the geological conditions based on the fusion data. And during the data fusion process, through optimization processing, the fused data can effectively reduce noise interference and improve the signal-to-noise ratio, and finally obtain the optimal fused data, so as to more accurately detect underground structure information and effectively ensure the accuracy of the detection result. Brief Description of the Drawings

[0028] Figure 1 It is a schematic structural diagram of the seismic and transient electromagnetic joint detection and acquisition device in the present invention;

[0029] Figure 2 It is a manual operation flow chart of the present invention;

[0030] Figure 3 It is a seismic and transient electromagnetic joint detection and processing flow chart of the present invention;

[0031] Figure 4 It is a schematic diagram of the membership degree function image of the formation wave velocity in the present invention;

[0032] Figure 5 It is a schematic diagram of the membership degree function image of the transient electromagnetic full-period apparent resistivity in the present invention;

[0033] Figure 6 It is a schematic diagram of the detection result of the present invention. Detailed Embodiment

[0034] The present invention will be further described below.

[0035] As Figure 2 and 3 shown, the specific steps of the present invention are as follows:

[0036] Step 1: Assemble the seismic and transient electromagnetic joint detection and acquisition device: As Figure 1As shown in the figure, the detection and acquisition device includes an array transient electromagnetic coal thickness detection device, an array seismic coal thickness detection device, a T-shaped support, a seismic transient acquisition station, and a control and analysis terminal; the array transient electromagnetic coal thickness detection device consists of 8 three-component transceiver integrated electromagnetic coils; the array seismic coal thickness detection device consists of 8 three-component piezoelectric sensors and 1 electromagnetic exciter, where the electromagnetic exciter is the seismic source; when performing seismic detection, the 8 three-component piezoelectric sensors and the electromagnetic exciter are detachably connected to the upper part of the T-shaped support, and the 8 three-component piezoelectric sensors and the electromagnetic exciter are arranged in a straight line at equal intervals, and the electromagnetic exciter is located in the middle of the multiple three-component piezoelectric sensors; the distance between each of the 8 three-component piezoelectric sensors and the electromagnetic exciter is 0.2 m. Each three-component piezoelectric sensor and the electromagnetic exciter are connected to the seismic transient acquisition station; when performing transient electromagnetic detection, the 8 three-component electromagnetic coils are detachably connected to the upper part of the T-shaped support, and the installation positions of the 8 three-component electromagnetic coils correspond one by one to the installation positions of each three-component piezoelectric sensor and the electromagnetic exciter; and each three-component electromagnetic coil is connected to the seismic transient acquisition station; where the seismic transient acquisition station has functions of power supply, acquisition, storage, and transmission, and the control and analysis terminal has functions of parameter setting, control, data processing, and display. The three-component electromagnetic coils are all transceiver integrated coils;

[0037] Step 2: Establish an observation system for detecting the remaining coal thickness of the roof: Taking the roadway running direction as the X direction, the roadway horizontal plane direction as the Y direction, and the detection depth direction as the Z direction, establish an observation system for detecting the remaining coal thickness of the roof;

[0038] Step 3: Install the seismic and transient electromagnetic combined detection and acquisition device: First, determine the position where the coal thickness needs to be detected, then install the 8 three-component piezoelectric sensors and the electromagnetic exciter, and then lift the T-shaped support until each three-component piezoelectric sensor and the electromagnetic exciter are tightly coupled with the coal wall at the top of the roadway, and the arrangement direction of the 8 three-component piezoelectric sensors and the electromagnetic exciter is along the roadway running direction, and the detection direction of the 8 three-component piezoelectric sensors is towards the top coal seam. At this time, the installation work is completed;

[0039] Step 4. Collect seismic data and transient electromagnetic data: When conducting seismic exploration, the control and analysis terminal controls the electromagnetic vibrator to start through the seismic transient acquisition station and excite seismic waves at its position. Eight three-component piezoelectric sensors receive the seismic wave data reflected from the coal seam, with a total of 24 seismic records, and transmit the seismic wave data to the control and analysis terminal through the seismic transient acquisition station; after completing the seismic data acquisition, lower the T-shaped support, disassemble the eight three-component piezoelectric sensors and the electromagnetic vibrator from the T-shaped support, then install the eight three-component electromagnetic coils one by one at the installation positions of the respective three-component piezoelectric sensors. After completion, raise the T-shaped support again until each three-component electromagnetic coil is tightly coupled with the coal wall at the top of the roadway. At this time, the layout work for transient electromagnetic exploration is completed. When conducting transient electromagnetic exploration, the control and analysis terminal controls one of the three-component electromagnetic coils to excite transient electromagnetic signals through the seismic transient acquisition station, and then the eight three-component electromagnetic coils respectively receive the transient electromagnetic data reflected from the coal seam, with a total of 24 transient electromagnetic records, and transmit the transient electromagnetic data to the control and analysis terminal through the seismic transient acquisition station; complete the process of collecting seismic data and transient electromagnetic data;

[0040] Step 5. Preliminarily process the seismic data: Taking the position of the electromagnetic vibrator as the origin, establish a plane XZ coordinate system. The X and Z directions of the XZ coordinate system are consistent with the observation system. Divide the coordinate system with a square grid with a side length of 0.1 m, and put the acquired seismic data into the above XZ coordinate system. According to the first seismic signal times t1, t2... t8 received by each three-component piezoelectric sensor in the grid and combined with its grid coordinate position, calculate the seismic apparent velocity V s1 、V s2 ……V s8 ; The specific calculation process of the seismic apparent velocity is as follows:

[0041] Assume that the coordinates of two adjacent piezoelectric sensors are (x1, y1, z1) and (x2, y2, z2) respectively, and their corresponding seismic apparent velocities are x, y, z are the coordinates of the grid where the piezoelectric sensor is located, and t is the first seismic signal time recorded by each piezoelectric sensor.

[0042] Then, conduct polarization analysis according to the seismic data of different components, and convert each seismic apparent velocity V s into the formation wave velocity V d1 V d2 ……V d8 in the main polarization direction. Finally, perform fuzzy processing on the formation wave velocity V d and optimize the membership function of the formation wave velocity. The optimization is specifically as follows: Due to the complexity and heterogeneity of the formation, the formation wave velocity V dIt usually does not show a simple linear relationship. Therefore, it is necessary to optimize the conventional membership function to handle this non-linear relationship and more accurately predict the distribution of the formation wave velocity V. d As Figure 4 shown, the hyperbolic tangent function is selected to optimize the membership function to describe the formation wave velocity V with non-linear growth and saturation characteristics. d The data, and the specific expression is:

[0043]

[0044] where A and B are the minimum and maximum values of the membership function, which are 0 and 1 respectively in this example. C is the parameter of the function change rate, which is 0.002 in this example; D is the translation amount of the function, and D up and D down represent the translation amounts of the rising section and the falling section respectively, which are 0 and 2000 respectively in this example.

[0045] Finally, the optimized seismic data fuzzy set A1 = f1(V d ) is obtained, where f1 is the membership function of the formation wave velocity;

[0046] Step 6: Preliminarily process the transient electromagnetic data: Convert the voltage signals collected for each component in the collected transient electromagnetic data into the full-period apparent resistivity ρ s1 , ρ s2 ... ρ s8 , and perform fuzzy processing on the full-period apparent resistivity data ρ s of the transient electromagnetic to obtain the transient electromagnetic data fuzzy set A2 = f2(ρ s ), where f2 is the membership function of the full-period apparent resistivity ρ s of the transient electromagnetic; Since the conventional coal thickness is basically in the early stage of electromagnetic wave attenuation for the transient electromagnetic, and the influence of small structures such as the formation structure during this period is relatively small. As Figure 5 shown, therefore, the conventional triangular membership function can be used to achieve the fuzzy processing of the transient electromagnetic data;

[0047] Step 7: Fuse the seismic data and the transient electromagnetic data: Construct a fuzzy logic rule base containing 2 rules. The rules in the fuzzy logic rule base need to be set in combination with the on-site formation conditions. Taking the overlying rock layer of the roof coal seam as a dense sandstone layer as an example, in the logic rule base: If A1 is high, then A2 is low; If A1 is low, then A2 is high. Where A1 and A2 are the fuzzy sets of the formation wave velocity V d and the full-period apparent resistivity ρ s respectively; that is, V d is A 1i , and ρ s is A 2j, where \(i\) represents the membership degree of the formation wave velocity, \(j\) represents the membership degree of the transient electromagnetic full-period apparent resistivity data, and \(w\) i , \(w\) j represent the corresponding weights; input the fuzzified data of Step Five and Step Six into the rule base for inference to obtain the fuzzy output set \(B\) n = min{\(A\) 1i , \(A\) 2j}, where \(i = 1, 2, \cdots, 8\), \(j = 1, 2, \cdots, 8\), and \(n\) represents the membership degree of the \(n\)th output; the fuzzy output set \(B\) n represents the membership degree of the fuzzy set of the inference result, which is composed of the seismic data fuzzy set \(A_1\), the transient electromagnetic data fuzzy set \(A_2\), and their respective membership degree values, and is obtained according to the fuzzy intersection operation. Convert the fuzzy output set \(B\) n into the actual weight value \(w\) n , \(w\) n = ∑\((i, j)B\) n * \(w(i, j)\) / ∑\((i, j)B\) n . Since small structures are contained in the coal seam, resulting in local anomalies in the formation velocity, the lithology and water content of the overlying strata of the coal seam will interfere with the detection data of transient electromagnetic methods. Therefore, when using the two methods alone, it will be difficult to overcome the bottleneck of the method itself and improve the accuracy. To meet the needs of detecting coal thickness under various geological conditions as much as possible, it is necessary to make adaptive improvements to the data fusion of the two, so as to automatically adjust the weight and membership degree parameters to match different geological situations. The specific optimization objective function \(T\) is:

[0048] \(T = w\) i * RMSE(\(V\) R , \(V\) d , \(f_1\)) + \(w\) j * RMSE(\(V\) R , \(ρ\) s , \(f_2\))

[0049] where \(V\) R is the fused data, \(V\) d is the formation wave velocity data, \(f_1\) is the membership function of the formation wave velocity, \(w\) i is the weight corresponding to the formation wave velocity, \(ρ\) s is the transient electromagnetic full-period apparent resistivity data, \(f_2\) is the membership function of the transient electromagnetic full-period apparent resistivity \(ρ\) s , \(w\) j is the weight corresponding to the transient electromagnetic full-period apparent resistivity, and RMSE represents the root mean square error, specifically:

[0050] RMSE(\(V\) R , \(V\) d , \(f_1\)) = (\(Σ(V\) Rk - \(V\) dk )^2 / k)^0.5

[0051] where V Rk is the k-th data point after fusion, V dk is the k-th formation wave velocity data, and k is the total number of data points. Based on this optimization objective function T, the optimal weights w i , w j and the membership function parameters A, B, C, D up , D down are found by means of an optimization algorithm, such that the objective function reaches the minimum value. Finally, weighted averaging is performed according to the optimized weight values corresponding to each data source to obtain the fused data; the data is fused by adaptively adjusting the weights and membership parameters, thereby ensuring the reliability and adaptability of this fusion algorithm.

[0052] Step Eight: Determine the remaining coal thickness: Use the fused data as formation parameters for conventional data inversion, and obtain the remaining coal thickness of the roof at the current position according to the inversion result. As Figure 6 shown, repeat the above steps for each subsequent position to obtain the remaining coal thickness of the roof at different positions.

[0053] The above seismic transient acquisition station, electromagnetic vibrator, three-component piezoelectric sensor, three-component electromagnetic coil, and control analysis terminal are all existing devices or components and can be obtained through market purchase.

[0054] The above are only the preferred embodiments of the present invention. It should be noted that: for those of ordinary skill in the art in this technical field, without departing from the principle of the present invention, several improvements and refinements can still be made, and these improvements and refinements should also be regarded as the protection scope of the present invention.

Claims

1. A method for jointly detecting the remaining coal thickness of the roof by seismic and transient electromagnetic methods, characterized in that, The specific steps are as follows: Step 1: Assemble the seismic and transient electromagnetic joint detection acquisition device: The detection acquisition device includes a T-shaped bracket, a seismic transient acquisition station, an electromagnetic exciter, multiple three-component piezoelectric sensors, multiple three-component electromagnetic coils, and a control and analysis terminal. The seismic transient acquisition station is connected to the control and analysis terminal. When conducting seismic detection, multiple three-component piezoelectric sensors and the electromagnetic exciter are detachably installed on the upper part of the T-shaped bracket. The multiple three-component piezoelectric sensors and the electromagnetic exciter are arranged in a straight line at equal intervals, and the electromagnetic exciter is located in the middle of the multiple three-component piezoelectric sensors. Each three-component piezoelectric sensor and the electromagnetic exciter are connected to the seismic transient acquisition station. When conducting transient electromagnetic detection, multiple three-component electromagnetic coils are detachably installed on the upper part of the T-shaped bracket. The installation positions of the multiple three-component electromagnetic coils correspond one by one to the installation positions of each three-component piezoelectric sensor. And each three-component electromagnetic coil is connected to the seismic transient acquisition station; Step 2: Establish an observation system for detecting the remaining coal thickness of the roof: Taking the roadway alignment direction as the X direction, the roadway horizontal plane direction as the Y direction, and the detection depth direction as the Z direction, establish an observation system for detecting the remaining coal thickness of the roof; Step 3: Layout the seismic and transient electromagnetic joint detection acquisition device: First, determine the position where the coal thickness needs to be detected. Then, install multiple three-component piezoelectric sensors and the electromagnetic exciter first. Next, lift the T-shaped bracket until each three-component piezoelectric sensor and the electromagnetic exciter are tightly coupled with the coal wall at the top of the roadway. And the arrangement direction of the multiple three-component piezoelectric sensors and the electromagnetic exciter is along the roadway alignment direction, and the detection direction of the multiple three-component piezoelectric sensors is towards the top coal seam. At this time, the layout work is completed; Step 4: Collect seismic data and transient electromagnetic data: When conducting seismic detection, the control and analysis terminal enables the electromagnetic exciter to start through the seismic transient acquisition station and excite seismic waves at its position. Each three-component piezoelectric sensor receives the seismic wave data reflected in the coal seam and transmits the seismic wave data to the control and analysis terminal through the seismic transient acquisition station; After the seismic data collection is completed, lower the T-shaped bracket, disassemble each three-component piezoelectric sensor and the electromagnetic exciter from the T-shaped bracket. Then, install multiple three-component electromagnetic coils respectively at the installation positions of each three-component piezoelectric sensor one by one. After completion, lift the T-shaped bracket again until each three-component electromagnetic coil is tightly coupled with the coal wall at the top of the roadway. At this time, the layout work for transient electromagnetic detection is completed. When conducting transient electromagnetic detection, the control and analysis terminal enables one of the three-component electromagnetic coils to excite transient electromagnetic signals through the seismic transient acquisition station. Then, all three-component electromagnetic coils respectively receive the transient electromagnetic data reflected in the coal seam and transmit the transient electromagnetic data to the control and analysis terminal through the seismic transient acquisition station; Complete the process of collecting seismic data and transient electromagnetic data; Step 5. Preliminary processing of seismic data: Taking the position of the electromagnetic shaker as the origin, establish a plane XZ coordinate system. The X and Z directions of the XZ coordinate system are consistent with the observation system. Divide the coordinate system with a square grid with a side length of 0.1 m, and put the acquired seismic data into the above XZ coordinate system. According to the time t of the first seismic signal received by each grid where the three-component piezoelectric sensors are located, combined with its grid coordinate position, calculate the seismic apparent velocity V received by each three-component piezoelectric sensor. s ; Then, perform polarization analysis according to the seismic data of different components, and convert each seismic apparent velocity V s into the formation wave velocity V in the main polarization direction d . Finally, perform fuzzification processing on the formation wave velocity V d , and optimize the membership function of the formation wave velocity to finally obtain the optimized seismic data fuzzy set A1 = f1(V d ), where f1 is the membership function of the formation wave velocity. Step 6. Preliminary processing of transient electromagnetic data: Convert the voltage signals collected for each component in the collected transient electromagnetic data into the transient electromagnetic full-period apparent resistivity ρ s , and perform fuzzification processing on the transient electromagnetic full-period apparent resistivity ρ s to obtain the transient electromagnetic data fuzzy set A2 = f2(ρ s ), where f2 is the membership function of the transient electromagnetic full-period apparent resistivity ρ s ; Step 7: Fuse seismic data and transient electromagnetic data: Construct a fuzzy logic rule base with 2 rules, input the data fuzzified in Step 5 and Step 6 into the rule base, and perform reasoning to obtain a fuzzy output set B n = min{A 1i , A 2j}, where n represents the membership degree of the nth output; Convert the fuzzy output set B n into an actual weight value w n , w n = ∑(i,j)B n * w(i,j) / ∑(i,j)B n , and make adaptive improvements to the fusion of seismic data and transient electromagnetic data. The specific objective function T of the adaptive improvement is as follows: T = w i *RMSE(V R , V d , f1) + w j *RMSE(V R , ρ s , f2) Among them, V R is the fused data, V d is the formation wave velocity data, f1 is the membership function of the formation wave velocity, w i is the weight corresponding to the formation wave velocity, ρ s is the apparent resistivity of the entire period of transient electromagnetic method, f2 is the membership function of the apparent resistivity ρ s of the entire period of transient electromagnetic method, w j is the weight corresponding to the apparent resistivity of the entire period of transient electromagnetic method, RMSE represents the root mean square error, specifically: RMSE(V R ,V d ,f1) = (Σ(V Rk - V dk )^2 / k)^0.5 Among which V Rk is the k-th data point after fusion, V dk is the k-th formation wave velocity data, and k is the total number of data points; based on this objective function T, the optimal weights w i , w j and the membership function parameters A, B, C, D up , D down are found by means of an optimization algorithm to minimize the objective function. Finally, weighted averaging is performed according to the optimized weight values corresponding to each data source to obtain the fused data; Step 8: Determine the remaining coal thickness: Use the fused data as formation parameters for conventional data inversion, and obtain the remaining coal thickness of the roof according to the inversion result.

2. The method for jointly detecting the remaining coal thickness of the roof by seismic and transient electromagnetic methods according to claim 1, characterized in that, The spacing between the multiple three-component piezoelectric sensors and the electromagnetic exciters is all 0.2 m.

3. The method for jointly detecting the remaining coal thickness of the roof by seismic and transient electromagnetic methods according to claim 1, characterized in that, The optimization in the fifth step is specifically as follows: The hyperbolic tangent function is selected to optimize the membership function for describing the formation wave velocity V with nonlinear growth and saturation characteristics d data, and the specific expression is: where A and B are the minimum and maximum values of the membership function respectively, C is the parameter of the function change rate, D is the translation amount of the function, D up and D down represent the translation amounts of the rising and falling segments respectively.

4. The method for jointly detecting the remaining coal thickness of the roof by seismic and transient electromagnetic methods according to claim 1, characterized in that, The calculation formula for the seismic apparent velocity in Step 5 is as follows: Where (x1, y1, z1) and (x2, y2, z2) are the coordinates of two adjacent piezoelectric sensors respectively; t1 and t2 are the times of the first seismic signals recorded by the two adjacent piezoelectric sensors respectively.

5. The method for jointly detecting the remaining coal thickness of the roof by seismic and transient electromagnetic methods according to claim 1, characterized in that, The rules in the fuzzy logic rule base in Step 7 need to be set in combination with the on-site formation conditions.

6. The method for jointly detecting the remaining coal thickness of the roof by seismic and transient electromagnetic methods according to claim 1, characterized in that, The fuzzy output set B n represents the membership degree of the fuzzy set of the inference result, which is composed of the seismic data fuzzy set A1, the transient electromagnetic data fuzzy set A2 and their respective membership degree values, and is obtained according to the fuzzy intersection operation.

Images