Coal mining machine self-adaptive height adjusting system and method based on coal-rock interface fine identification

Abstract

The invention discloses a coal cutter self-adaptive height adjustment system and method based on coal rock interface fine identification, and the system comprises a coal seam thickness detection module, a cutting state detection module and a coal cutter height adjustment control module. The coal seam thickness detection module obtains a single-tool coal-rock interface under a mine absolute coordinate system according to the electromagnetic wave signal reflected by the coal-rock interface, the tilt angle data of each hydraulic support and the pushing slip amount, and stores the single-tool coal-rock interface to a historical-tool coal-rock interface database; the cutting state detection module collects cutting sound wave signals, discriminates the real-time cutting state of the roller of the coal mining machine according to the cutting sound wave signals and a historical cutting number coal rock interface, and carries out refined identification and local correction on the coal rock interface; and the coal cutter height adjustment control module predicts a current cutter number coal rock interface based on historical cutter number coal rock interface information, plans a target cutting path of a coal cutter roller, and performs coal cutter height adjustment control. The method has the characteristics of high coal mining rate, good economical efficiency, high precision and reliable system.

Description

Technical Field

[0001] The invention relates to a coal mining machine height adjustment system and method, and in particular to a coal mining machine self-adaptive height adjustment system and method based on refined recognition of coal-rock interface. Background Art

[0002] Due to the low visibility and high noise of the working face, the changes in the coal-rock interface are complex, and it is very difficult to identify the coal-rock interface. The coal mining machine is prone to cutting the top and bottom plate rocks, which increases the wear of the drum pick and affects the life of the pick; the cut coal is mixed with gangue, which reduces the quality of the coal; for high-gas mines, it is easy to cause explosion accidents; insufficient cutting will cause waste of coal resources, and a high recovery rate cannot be obtained, affecting economic benefits. At present, the main ways to adjust the height of the coal mining machine are manual adjustment and memory cutting, which are overly dependent on the operating level of the staff, have weak adaptive ability, and cannot adapt to working faces with complex geological conditions. Summary of the invention

[0003] Purpose of the invention: One purpose of the present invention is to provide a coal mining machine adaptive height adjustment system for coal-rock interface identification based on electromagnetic / acoustic wave refined deep fusion.

[0004] Another object of the present invention is to provide a method for adaptively adjusting the height of a coal mining machine based on the above system.

[0005] Technical solution: The coal mining machine adaptive height adjustment system based on refined identification of coal-rock interface described in the present invention works in the comprehensive mining working face according to the working characteristics of the three machines, including a coal seam thickness detection module, a cutting state detection module and a coal mining machine height adjustment control module. The coal seam thickness detection module obtains the single-knife coal-rock interface in the absolute coordinate system of the mine according to the electromagnetic wave signal reflected from the coal-rock interface, the swing angle data of each hydraulic support and the sliding amount, and stores it in the historical knife coal-rock interface database; the cutting state detection module collects the cutting sound wave signal, and determines the real-time cutting state of the coal mining machine drum according to the cutting sound wave signal and the historical knife coal-rock interface, and performs refined identification and local correction of the coal-rock interface; the coal mining machine height adjustment control module predicts the current knife coal-rock interface based on the historical knife coal-rock interface information, plans the target cutting path of the coal mining machine drum, and performs coal mining machine height adjustment control.

[0006] Optionally, the coal seam thickness detection module includes a low-frequency ultra-wideband pulse radar, a support inclination sensor, a linear sensor and an embedded system I, wherein the low-frequency ultra-wideband pulse radar transmits an electromagnetic wave signal to the coal-rock interface, and receives the electromagnetic wave signal reflected from the coal-rock interface; the support inclination sensor and the linear sensor respectively collect the swing angle data and the displacement slippage of each hydraulic support; the embedded system I estimates the remaining coal seam thickness value located directly above the radar transceiver antenna at the last cut based on the electromagnetic wave signal reflected from the coal-rock interface, calculates the vertical height of the top of each hydraulic support roof based on the swing angle data of each hydraulic support, and obtains a single-cut coal-rock interface including M sampling points in the absolute coordinate system of the mine in combination with the displacement slippage of a single hydraulic support.

[0007] Optionally, the low-frequency ultra-wideband pulse radar is installed on the side plate of the hydraulic support, and a radar is installed on every N hydraulic supports. The radars are arranged in an equally spaced linear array. The low-frequency ultra-wideband pulse radar works in an air-coupled manner, and the transmitting antenna radiates electromagnetic wave signals directly upward, and the receiving antenna receives electromagnetic wave signals reflected from the air-coal seam and the coal-rock layer.

[0008] Optionally, the cutting state detection module includes a sound sensor and an embedded system II, the sound sensor is used to detect the cutting sound wave signal when the coal mining machine drum cuts coal and rock; the embedded system II determines the real-time cutting state of the coal mining machine drum according to the cutting sound wave signal and the historical coal-rock interface, and performs refined identification and local correction of the coal-rock interface.

[0009] Optionally, the real-time cutting status of the coal mining machine drum includes the cutting status of coal seams, faults, interlayers and roof and floor plates.

[0010] Optionally, the coal mining machine height adjustment control module includes a height adjustment cylinder displacement sensor, a strapdown inertial navigation module and an embedded system III, wherein the height adjustment cylinder displacement sensor is used to detect the actual displacement of the coal mining machine height adjustment cylinder; the strapdown inertial navigation module is used to collect the displacement of the coal mining machine body; the embedded system III predicts the current cut coal-rock interface based on historical cut coal-rock interface information, plans the target cutting path of the coal mining machine drum, and performs coal mining machine height adjustment control.

[0011] The adaptive height adjustment method using the above system includes the following steps:

[0012] (1) The coal mining machine travels on the scraper conveyor and performs the first cut manually before adaptive cutting;

[0013] (2) The hydraulic support moves smoothly, pushing the scraper conveyor and coal mining machine forward radially along the working face;

[0014] (3) The low-frequency ultra-wideband pulse radar, the support inclination sensor and the linear sensor work. The low-frequency ultra-wideband pulse radar transmits electromagnetic wave signals to the coal-rock interface in real time, and receives the electromagnetic wave signals reflected from the coal-rock interface, i.e., the electromagnetic wave receiving signals; the support inclination sensor and the linear sensor respectively collect the swing angle data and the displacement slippage of each hydraulic support in real time; the embedded system I calculates the single-knife coal-rock interface including M sampling points in the absolute coordinate system of the mine according to the electromagnetic wave receiving signals, the swing angle data and the displacement slippage of each hydraulic support;

[0015] (4) Embedded system III predicts the coal-rock interface of the current cut based on the LSTM network according to the coal-rock interface information of the historical cuts, and performs feature extraction to obtain the electromagnetic wave feature vector;

[0016] (5) Embedded system III considers equipment limitations, coal quality, and continuity requirements, takes the maximum recovery rate as the goal, uses the constrained optimization algorithm to perform constraint optimization of the coal-rock interface of the current cut, and calculates the target cutting path of the current cut;

[0017] (6) The height adjustment cylinder displacement sensor and the strapdown inertial navigation module work. The height adjustment cylinder displacement sensor collects the actual displacement of the coal mining machine height adjustment cylinder in real time, and the strapdown inertial navigation module collects the displacement of the fuselage in real time. The embedded system III converts the coal mining machine drum height adjustment problem into a height adjustment cylinder trajectory tracking problem based on the inverse kinematics model of the coal mining machine height adjustment system and the fuselage displacement. The target displacement of the height adjustment cylinder is calculated according to the target cutting trajectory, and compared with the actual displacement of the height adjustment cylinder, the height adjustment cylinder is controlled to move according to the target displacement, thereby controlling the coal mining machine to use the target cutting trajectory as the target cutting path of the coal mining machine drum to perform the current cut work.

[0018] (7) At the same time, the sound sensor works, the sound sensor collects the cutting sound wave signal in real time, uses the strapdown inertial navigation module to obtain the displacement of the fuselage in real time, and the embedded system II determines the real-time cutting state of the coal mining machine drum, adjusts the drum height in real time according to the real-time cutting state of the coal mining machine drum, and partially corrects the coal-rock interface information. After the single-knife cutting is completed, the current knife-step digital coal-rock interface is generated, and the historical knife-step coal-rock interface is stored and recorded;

[0019] (8) Repeat steps (2) to (7) until the adaptive cutting of the fully mechanized mining face is completed.

[0020] Furthermore, the calculation of the single-knife coal-rock interface in step (3) includes the following steps: calculating the covariance matrix of the electromagnetic wave signal reflected from the coal-rock interface, calculating the round-trip time delay between air and coal seam and coal and rock layer using the MUSIC method, and calculating the air layer thickness value and the remaining coal seam thickness value according to the propagation speed of electromagnetic waves in the air and coal seam; adding the vertical height of the top of the hydraulic support roof, the air layer thickness value, and the remaining coal seam thickness value to obtain the coal-rock interface point height value; and obtaining the single-knife coal-rock interface information according to the detection results of the low-frequency ultra-wideband pulse radar array.

[0021] Furthermore, judging the real-time cutting state of the coal mining machine drum in step (7) includes the following steps: extracting the multi-layer wavelet packet components of the cutting sound wave signal, adaptively extracting the sound wave feature vector that can characterize the cutting state based on the Transformer module and the variational encoder, connecting the electromagnetic wave feature vector, the body displacement, and the sound wave feature vector, and inputting them into the fully connected layer to judge the cutting state of the coal mining machine drum in real time.

[0022] Furthermore, in step (7), when the real-time cutting state of the coal mining machine drum is judged to be cutting coal seams and interlaced with gangue, the coal mining machine drum moves along the drum target trajectory; when it is judged to be the top plate or the bottom plate or the fault, the height adjustment cylinder is controlled to be retracted, that is, the upper drum moves downward and the lower drum moves upward, until the real-time cutting state of the drum changes to cutting coal seams and interlaced with gangue. Beneficial effect: Compared with the prior art, the significant technical effect of the present invention is that it can solve the problems of low precision and poor working environment of the manual recognition interface, and low control precision and poor applicability of the traditional control method, so that the coal mining machine can cut the coal seam according to the actual working conditions, so as to achieve the purposes of improving efficiency, reducing the amount of coal interlaced with gangue, increasing the life of the cutting teeth, increasing the recovery rate and protecting the health of workers, etc., which has both economic and social benefits. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 It is a system structure block diagram of the present invention;

[0024] Figure 2 The figure is a flow chart of the method of the present invention. DETAILED DESCRIPTION

[0025] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments.

[0026] like Figure 1As shown, the system of the present invention includes: a coal seam thickness detection module, a cutting state detection module and a coal mining machine height control module. The coal seam thickness detection module transmits an electromagnetic wave signal (called an electromagnetic wave transmission signal) to the coal-rock interface, and receives the electromagnetic wave signal reflected back from the coal-rock interface (called an electromagnetic wave receiving signal). According to the electromagnetic wave receiving signal, the swing angle data of each hydraulic support and the sliding amount, the single-knife coal-rock interface information including M sampling points in the absolute coordinate system of the mine is obtained, and stored in the historical knife coal-rock interface database; the cutting state detection module collects the cutting sound wave signal, based on the electromagnetic wave receiving signal, the swing angle data of each hydraulic support and the sliding amount. According to the cutting sound wave signal and the coal-rock interface of the historical cut, the real-time cutting state of the coal-cutting machine drum is judged, and the coal-rock interface is finely identified and partially corrected; before the second cut of the coal-cutting machine is completed, the coal-rock interface of the historical cut is the single-cut coal-rock interface obtained by the coal seam thickness detection module, and in the subsequent work, the coal-rock interface of the historical cut is the coal-rock interface information of the historical cut obtained by the coal seam thickness detection module and the cutting state detection module; the coal-cutting machine height adjustment control module predicts the coal-rock interface of the current cut based on the coal-rock interface information of the historical cut, plans the target cutting path of the coal-cutting machine drum, and performs the coal-cutting machine height adjustment control. The present invention has the characteristics of high coal mining rate, good economy, high precision and reliable system.

[0027] The coal seam thickness detection module includes a low-frequency ultra-wideband pulse radar, a support inclination sensor, a linear sensor and an embedded system I; the low-frequency ultra-wideband pulse radar is fixedly installed on the side plate of the hydraulic support, transmits electromagnetic wave signals to the coal-rock interface, and receives electromagnetic wave signals reflected from the coal-rock interface (i.e., electromagnetic wave receiving signals); the support inclination sensor and the linear sensor are installed on each hydraulic support, and the swing angle data and the displacement slippage of each hydraulic support are respectively collected; the embedded system I is installed on the body of the coal mining machine, and the remaining coal seam thickness value located directly above the radar transceiver antenna of the previous cut is estimated according to the electromagnetic wave receiving signal, and the vertical height of the top of the hydraulic support roof is calculated according to the swing angle data of each hydraulic support, and the single-cut coal-rock interface including M sampling points in the absolute coordinate system of the mine is obtained in combination with the displacement slippage of a single hydraulic support.

[0028] The low-frequency ultra-wideband pulse radar is installed on the side panels of the hydraulic support, and a radar is installed on every N hydraulic supports, and the radars are arranged in an equally spaced linear array. The low-frequency ultra-wideband pulse radar works in an air-coupled manner, the transmitting antenna radiates electromagnetic wave signals directly upward, and the receiving antenna receives electromagnetic wave signals reflected from the air-coal seam and the coal-rock layer.

[0029] The cutting state detection module includes a sound sensor and an embedded system II; the sound sensor is installed in the coal mining machine drum to detect the cutting sound wave signal when the coal mining machine drum cuts coal and rock; the embedded system II is installed on the coal mining machine body to determine the real-time cutting state of the coal mining machine drum according to the cutting sound wave signal, electromagnetic wave characteristic vector, and body displacement, and perform refined recognition and local correction of the coal-rock interface.

[0030] When the cutting state detection module performs coal-rock interface refinement and local correction, the cutting acoustic wave signal x s (t) After multi-layer wavelet packet decomposition, for the J-th layer decomposition, the wavelet packet function is expressed as:

[0031]

[0032] Among them, a J,n is the wavelet packet coefficient of the Jth layer decomposition, ψ J,n (t) is the time shift and scale transformation of the mother wavelet function of the Jth layer decomposition, n = 1, 2, ..., 2 J is the number of coefficients in this frequency band. Then, the multi-layer wavelet packet components of the cut acoustic signal are extracted Input into the Transformer-VAE model, adaptively extract the acoustic wave feature vector W1′ that can characterize the cut state. The loss function L1 of the Transformer-VAE model is defined as:

[0033]

[0034] where KL(·) is the KL divergence, is a Gaussian distribution, μ is the mean, and σ is the variance. Finally, the electromagnetic wave feature vector I1, the acoustic wave feature vector I2, and the fuselage displacement S are connected to obtain the feature vector I3, which is input into the fully connected layer to determine the drum cutting state in real time. The real-time cutting state of the drum detected by the cutting state detection module includes the cutting of coal seams, faults, interlayers, roof and floor plates, etc.

[0035] The shearer height control module includes a height cylinder displacement sensor, a strapdown inertial navigation module and an embedded system III; the height cylinder displacement sensor is installed on the shearer rocker arm to detect the actual displacement of the height cylinder; the strapdown inertial navigation module is installed on the shearer body to collect the absolute position information of the shearer body; the embedded system III is installed on the shearer body to predict the current cut coal-rock interface based on the historical cut coal-rock interface information, and use the constrained optimization algorithm to plan the target cutting path of the shearer drum according to the current cut coal-rock interface, and perform shearer height control. Figure 2 As shown, the method for adaptively adjusting the height of a coal mining machine using the above system comprises the following steps:

[0036] (a) The coal mining machine travels on the scraper conveyor and performs the first cut by manual operation before adaptive cutting;

[0037] (b) The hydraulic support pushes the scraper conveyor and coal mining machine forward radially along the working face;

[0038] (c) The low-frequency ultra-wideband pulse radar, the support tilt sensor, and the linear sensor are working. The low-frequency ultra-wideband pulse radar transmits electromagnetic wave signals to the coal-rock interface in real time, and receives the electromagnetic wave signals reflected back from the coal-rock interface (i.e., electromagnetic wave receiving signals); the support tilt sensor and the linear sensor respectively collect the swing angle data and the sliding amount of each hydraulic support in real time; the embedded system I calculates the single-knife coal-rock interface including M sampling points in the absolute coordinate system of the mine;

[0039] (d) Embedded system III predicts the coal-rock interface of the current cut based on the LSTM network according to the coal-rock interface information of the historical cuts, and performs feature extraction to obtain the electromagnetic wave feature vector;

[0040] Furthermore, the method for predicting the coal-rock interface of the current cut is as follows: the coal-rock interface information of each cut is recorded as y t , where t represents the cutting number, and the coal-rock interface information of the historical cutting number [y1, y2,…, y T ] is input into the LSTM network to predict the coal-rock interface y of the next cut T+1 , the training label value is The loss function of the LSTM model Defined as:

[0041]

[0042] (e) Embedded system III considers equipment limitations, coal quality, and continuity requirements, and uses a constrained optimization algorithm to perform the next cut of the coal-rock interface y with the maximum recovery rate as the goal. T+1 Constrained optimization is performed to calculate the target cutting path h for the current cut;

[0043] Furthermore, the specific steps of using the constrained optimization algorithm to obtain the target cutting path h are as follows:

[0044] According to equipment limitations, coal quality, and continuity requirements, the target cutting path smoothness constraint q1(k), cutting trajectory endpoint position constraint q2(k), and rock cutting rate constraint q3(k) are established. Then, with the maximum recovery rate as the goal, the objective function RMSE is established:

[0045]

[0046] According to the objective function and constraints, the target cutting path is determined using the constrained optimization algorithm.

[0047] (f) The oil cylinder displacement sensor and the strapdown inertial navigation module work. The oil cylinder displacement sensor collects the actual displacement of the coal mining machine oil cylinder in real time, and the strapdown inertial navigation module collects the displacement of the fuselage in real time. The embedded system III converts the drum height adjustment problem into a cylinder height adjustment trajectory tracking problem according to the inverse kinematics model of the coal mining machine height adjustment system and the fuselage displacement. The target displacement of the oil cylinder is calculated according to the target cutting trajectory, and compared with the actual displacement of the oil cylinder, the oil cylinder is controlled to move according to the target displacement, thereby controlling the coal mining machine (1) to use the target cutting trajectory as the drum target cutting path to perform the current knife operation. Furthermore, the target displacement L of the oil cylinder is calculated according to the target cutting trajectory. C The expression is as follows:

[0048]

[0049] Among them, L I is the initial position of the given piston rod, L H is the cutting height of the target cutting trajectory, L L is the length of the large rocker arm, L S is the length of the small rocker arm, L J To increase the distance between the hinge point between the oil cylinder and the fuselage and the hinge point between the rocker arm and the fuselage, the angle between their connecting line and the positive direction of the y-axis is θ A ,θ B It is the angle between the large and small rocker arms of the coal mining machine.

[0050] (g) At the same time, the sound sensor works, the sound sensor collects the cutting sound wave signal in real time, uses the strapdown inertial navigation module to obtain the displacement of the fuselage in real time, and the embedded system II determines the real-time cutting state of the drum, adjusts the height of the drum in real time according to the cutting state, and partially corrects the coal-rock interface information. After the single-knife cutting is completed, the current knife-step digital coal-rock interface is generated, and the historical knife-step coal-rock interface is stored and recorded;

[0051] Furthermore, the strategy of adjusting the drum height in real time according to the cutting state and partially correcting the coal-rock interface information is as follows: when the real-time cutting state of the shearer drum is judged to be cutting the coal seam or sandwiching gangue, the shearer drum moves along the drum target trajectory; when it is judged to be the roof, floor, or fault, the height adjustment cylinder is controlled to be retracted, that is, the upper drum moves downward and the lower drum moves upward, until the real-time cutting state of the drum becomes cutting the coal seam or sandwiching gangue. At the same time, the coal-rock interface information is updated and corrected.

[0052] (h) Repeat steps (b) to (g) until the adaptive cutting of the fully mechanized mining face is completed.

[0053] When the real-time cutting state of the drum is judged to be cutting coal seams and interlacing gangue, the coal mining machine drum moves along the drum target trajectory. When it is judged to be the top plate, bottom plate or fault, the embedded system III controls the height adjustment cylinder to retract, that is, the upper drum moves downward and the lower drum moves upward, until the real-time cutting state of the drum becomes cutting coal seams and interlacing gangue.

[0054] The embedded system I calculates the single-knife coal-rock interface, including the following steps: calculating the covariance matrix of the electromagnetic wave receiving signal x, calculating the round-trip time delay between air and coal seam and coal and rock seam by using the MUSIC method, and calculating the air layer thickness value and the remaining coal seam thickness value according to the propagation speed of electromagnetic waves in the air and coal seam; adding the vertical height of the top of the hydraulic support roof, the air layer thickness value, and the remaining coal seam thickness value to obtain the height value of the coal-rock interface point above the radar; and obtaining the single-knife coal-rock interface information according to the electromagnetic wave receiving signal obtained by the low-frequency ultra-wideband pulse radar array.

[0055] The embedded system II judges the real-time cutting state of the drum including the following steps: extracting the multi-layer wavelet packet components of the cutting sound wave signal, adaptively extracting the sound wave feature vector that can characterize the cutting state based on the Transformer module and variational encoder (Transformer-VAE), connecting the electromagnetic wave feature vector, the body displacement, and the sound wave feature vector, and inputting them into the fully connected layer to judge the cutting state of the drum in real time.

Claims

1. A coal mining machine adaptive height adjustment system based on refined recognition of coal-rock interface, which works in a fully mechanized mining face according to the working characteristics of three machines, and is characterized by: It includes a coal seam thickness detection module, a cutting state detection module and a coal mining machine height adjustment control module. The coal seam thickness detection module obtains the single-knife coal-rock interface in the absolute coordinate system of the mine according to the electromagnetic wave signal reflected from the coal-rock interface, the swing angle data of each hydraulic support and the sliding amount, and stores it in the historical knife coal-rock interface database; the cutting state detection module collects the cutting sound wave signal, determines the real-time cutting state of the coal mining machine drum according to the cutting sound wave signal and the historical knife coal-rock interface, and performs refined identification and local correction of the coal-rock interface; the coal mining machine height adjustment control module predicts the current knife coal-rock interface based on the historical knife coal-rock interface information, plans the target cutting path of the coal mining machine drum, and performs coal mining machine height adjustment control.

2. The self-adaptive height adjustment system for coal mining machine based on refined identification of coal-rock interface according to claim 1 is characterized in that: The coal seam thickness detection module includes a low-frequency ultra-wideband pulse radar, a support inclination sensor, a linear sensor and an embedded system I. The low-frequency ultra-wideband pulse radar transmits electromagnetic wave signals to the coal-rock interface and receives electromagnetic wave signals reflected from the coal-rock interface; the support inclination sensor and the linear sensor respectively collect the swing angle data and the sliding amount of each hydraulic support; Embedded system I estimates the remaining coal seam thickness value located directly above the radar transceiver antenna in the previous cut based on the electromagnetic wave signal reflected from the coal-rock interface, calculates the vertical height of the top of each hydraulic support roof according to the swing angle data of each hydraulic support, and combines the slippage of a single hydraulic support to obtain a single-cut coal-rock interface including M sampling points in the absolute coordinate system of the mine.

3. The self-adaptive height adjustment system for coal mining machine based on refined identification of coal-rock interface according to claim 2 is characterized in that: The low-frequency ultra-wideband pulse radar is installed on the side plate of the hydraulic support. A radar is installed on every N hydraulic supports. The radars are arranged in an equally spaced linear array. The low-frequency ultra-wideband pulse radar works in an air-coupled manner. The transmitting antenna radiates electromagnetic wave signals directly upward, and the receiving antenna receives electromagnetic wave signals reflected from the air-coal seam and the coal-rock layer.

4. The self-adaptive height adjustment system for coal mining machine based on refined identification of coal-rock interface according to claim 1 is characterized in that: The cutting state detection module includes a sound sensor and an embedded system II, wherein the sound sensor is used to detect the cutting sound wave signal when the coal mining machine drum cuts coal and rock; Embedded system II determines the real-time cutting status of the shearer drum according to the cutting acoustic wave signal and the coal-rock interface of historical cuts, and performs refined identification and local correction of the coal-rock interface.

5. The self-adaptive height adjustment system for coal mining machine based on refined identification of coal-rock interface according to claim 4 is characterized in that: The real-time cutting status of the coal mining machine drum includes the cutting coal seam, fault, gangue and roof and floor status.

6. The self-adaptive height adjustment system for coal mining machine based on refined identification of coal-rock interface according to claim 1 is characterized in that: The coal mining machine height adjustment control module includes a height adjustment cylinder displacement sensor, a strapdown inertial navigation module and an embedded system III, wherein the height adjustment cylinder displacement sensor is used to detect the actual displacement of the coal mining machine height adjustment cylinder; The strapdown inertial navigation module is used to collect the displacement of the coal mining machine body; Embedded system III predicts the current cut coal-rock interface based on the historical cut coal-rock interface information, plans the target cutting path of the shearer drum, and performs height control of the shearer.

7. A method for adaptively adjusting the height of a coal mining machine based on refined identification of coal-rock interface, characterized in that: The following steps are involved: (1) The coal mining machine travels on the scraper conveyor and performs the first cut manually before adaptive cutting; (2) The hydraulic support moves smoothly, pushing the scraper conveyor and coal mining machine forward radially along the working face; (3) The low-frequency ultra-wideband pulse radar, the support tilt sensor and the linear sensor work. The low-frequency ultra-wideband pulse radar transmits electromagnetic wave signals to the coal-rock interface in real time, and receives the electromagnetic wave signals reflected back from the coal-rock interface, i.e., the electromagnetic wave receiving signals; the support tilt sensor and the linear sensor respectively collect the swing angle data and the sliding amount of each hydraulic support in real time; The embedded system I calculates the single-knife coal-rock interface including M sampling points in the absolute coordinate system of the mine according to the electromagnetic wave receiving signal, the swing angle data of each hydraulic support and the sliding amount; (4) Embedded system III predicts the coal-rock interface of the current cut based on the LSTM network according to the coal-rock interface information of the historical cuts, and performs feature extraction to obtain the electromagnetic wave feature vector; (5) Embedded system III considers equipment limitations, coal quality, and continuity requirements, takes the maximum recovery rate as the goal, uses the constrained optimization algorithm to perform constraint optimization of the coal-rock interface of the current cut, and calculates the target cutting path of the current cut; (6) The oil cylinder displacement sensor and the strapdown inertial navigation module work, the oil cylinder displacement sensor collects the actual displacement of the coal mining machine oil cylinder in real time, and the strapdown inertial navigation module collects the displacement of the fuselage in real time; Embedded system III converts the problem of raising the drum of the coal mining machine into the problem of tracking the trajectory of the raising cylinder according to the inverse kinematics model of the raising system of the coal mining machine and the displacement of the machine body. It calculates the target displacement of the raising cylinder according to the target cutting trajectory, compares it with the actual displacement of the raising cylinder, controls the raising cylinder to move according to the target displacement, and thus controls the coal mining machine to use the target cutting trajectory as the target cutting path of the coal mining drum to perform the current cutting operation. (7) At the same time, the sound sensor works, the sound sensor collects the cutting sound wave signal in real time, uses the strapdown inertial navigation module to obtain the displacement of the fuselage in real time, and the embedded system II determines the real-time cutting state of the coal mining machine drum, adjusts the drum height in real time according to the real-time cutting state of the coal mining machine drum, and partially corrects the coal-rock interface information. After the single-knife cutting is completed, the current knife-step digital coal-rock interface is generated, and the historical knife-step coal-rock interface is stored and recorded; (8) Repeat steps (2) to (7) until the adaptive cutting of the fully mechanized mining face is completed.

8. The method for adaptively adjusting the height of a coal mining machine based on refined identification of the coal-rock interface according to claim 7 is characterized in that: Calculating the single-knife coal-rock interface in step (3) includes the following steps: calculating the covariance matrix of the electromagnetic wave signal reflected from the coal-rock interface, calculating the round-trip time delay between air and coal seam and between coal and rock layer using the MUSIC method, and calculating the air layer thickness value and the remaining coal seam thickness value according to the propagation speed of electromagnetic waves in the air and coal seam; adding the vertical height of the top of the hydraulic support roof, the air layer thickness value and the remaining coal seam thickness value to obtain the coal-rock interface point height value; and obtaining the single-knife coal-rock interface information according to the detection results of the low-frequency ultra-wideband pulse radar array.

9. The method for adaptively adjusting the height of a coal mining machine based on refined identification of the coal-rock interface according to claim 7 is characterized in that: In step (7), judging the real-time cutting state of the coal mining machine drum includes the following steps: extracting the multi-layer wavelet packet components of the cutting sound wave signal, based on the Transformer module and the variational encoder, adaptively extracting the sound wave feature vector that can characterize the cutting state, connecting the electromagnetic wave feature vector, the body displacement, and the sound wave feature vector, and inputting them into the fully connected layer to judge the cutting state of the coal mining machine drum in real time.

10. The method for self-adapting height adjustment of coal mining machine based on refined identification of coal-rock interface according to claim 7, characterized in that: In step (7), when the real-time cutting state of the coal mining machine drum is judged to be cutting coal seams and interlacing gangue, the coal mining machine drum moves along the drum target trajectory; when it is judged to be the top plate, bottom plate or fault, the height adjustment cylinder is controlled to be retracted, that is, the upper drum moves downward and the lower drum moves upward, until the real-time cutting state of the drum becomes cutting coal seams and interlacing gangue.

Images