Automatic cutting depth control method for continuous mining of hard rock strata based on deep learning

Abstract

The present invention discloses an automatic cutting depth control method for continuous mining of hard rock strata based on deep learning. The control method includes the establishment and training of a deep neural network model, pre-drilling and fracturing operations and prediction of rock strata strength parameters, determination of a reasonable cutting depth control value and monitoring and control of continuous mining operations; while improving the cuttability of hard rock strata through pre-drilling and fracturing operations, predicting the rock strata strength parameters after fracturing according to a deep learning algorithm, and then determining a reasonable cutting depth value, which is beneficial to improving the mining efficiency of continuous mining equipment in a hard rock strata environment and reducing the consumption of cutting tools; then calculating the current continuous mining equipment operation strategy in real time according to real-time monitoring data and the trained deep neural network model, and then realizing real-time monitoring and control of cutting depth, cutting mechanism power and mining working surface flatness, which is beneficial to improving the control accuracy, safety and automation of continuous mining equipment.

Description

Technical Field

[0001] The present invention belongs to the technical field of neural network models and relates to an automatic cutting depth control method for continuous mining of hard rock strata based on deep learning. Background Art

[0002] The deep learning algorithm is a machine learning method based on artificial neural networks. Through multi-layer neural networks and large amounts of data training, it can automatically learn the features contained in the data and find the optimal expression of the task. In the field of deep learning, the core idea of ​​deep neural network model technology is to progressively extract features and abstractly express input data layer by layer through multiple hidden layers between the input layer and the output layer. With the significant improvement of computing power, deep neural network model technology has been widely used in various industrial scenarios, especially in mining operations where practical experience is very rich but theoretical results are relatively scarce.

[0003] A new type of continuous mining equipment is used in conjunction with a pre-drilling fracturing device. First, the drill tool of the pre-drilling fracturing device is used to drill a test hole and a vibration device is used to perform vibration fracturing, so that microcracks expand along the periphery of the drill hole in the hard rock stratum of the open-pit mine working face, thereby improving the cuttability of the hard rock stratum. Subsequently, the rock stratum is continuously cut by the cutting mechanism of the new type of continuous mining equipment, thereby realizing non-blasting, safe, environmentally friendly and economical continuous mining of hard rock strata. In view of the uneven distribution of hard rock strata, the nonlinearity of rock mechanical properties and the randomness of microcrack expansion, the strength characteristics of the rock stratum after pre-drilling fracturing are suitable for real-time prediction using deep neural network model technology.

[0004] The cutting depth is the depth that the continuous mining equipment cuts into the rock formation in one cutting cycle. During the operation in the hard rock environment, the cutting depth of the new continuous mining equipment is a major factor in determining the production capacity of the mining face. Insufficient cutting depth will lead to low mining efficiency, while excessive cutting depth will lead to reduced rock breaking efficiency and increased wear of the cutting mechanism. On the other hand, in the mining operation of the same working face, if the cutting depths at different positions are inconsistent, the flatness of the open-pit mine working face will be insufficient, which is not conducive to the control accuracy and automation of the continuous mining equipment. Therefore, it is necessary to combine the strength characteristics of the rock formation after pre-drilling and the material fatigue strength characteristics of the cutting mechanism of the continuous mining equipment to determine a reasonable cutting depth value, and determine the continuous mining equipment operation strategy in real time through the deep neural network model technology, so as to realize the control of continuous mining operation parameters including cutting depth, and improve the automation, intelligence, safety and efficiency of continuous mining operations in hard rock formations. Summary of the invention

[0005] In view of the shortcomings of the prior art, the present invention proposes an automatic cutting depth control method for continuous mining of hard rock strata based on deep learning, which is suitable for automatic control of cutting depth during operation of open-pit continuous mining equipment in hard rock environments, which is beneficial to improving mining efficiency and reducing the wear of the cutting mechanism, and realizing the control of cutting depth and automation, intelligence, safety and efficiency of continuous mining operations in hard rock strata.

[0006] The present application discloses a method for automatic cutting depth control in continuous mining of hard rock strata based on deep learning, which method comprises the following steps:

[0007] S100, establishing and training a deep neural network model: establishing an initial deep neural network model, including an input layer, a hidden layer and an output layer; the input layer is used to receive data, the hidden layer is used for data processing and feature extraction, and the output layer is used to generate prediction results;

[0008] The data received by the input layer include but are not limited to normalized pre-drilling technical parameters, mining environment state parameters and continuous mining equipment state parameters at different times; the prediction results generated by the output layer are rock formation strength parameters and continuous mining equipment control quantities;

[0009] Continuous mining data under hard rock conditions are obtained by screening historical data, and the continuous mining data include but are not limited to pre-drilling technical parameters, mining environment status parameters, continuous mining equipment status parameters, cutting depth parameters, rock formation strength parameters and continuous mining equipment control quantities at different times; the pre-drilling technical parameters include but are not limited to hole depth, hole diameter, drilling hole distribution spacing, drilling energy consumption, vibration fracturing times and original rock formation strength parameters obtained from geological survey data; the mining environment status parameters include but are not limited to the geometric shape of the rock formation working face; the continuous mining equipment status parameters include but are not limited to the continuous mining equipment position, cutting mechanism linear speed, cutting mechanism angular speed, cutting mechanism output power; the rock formation strength parameters include but are not limited to the average tensile strength of the rock formation after fracturing; the continuous mining equipment control quantities include but are not limited to the throttle control opening, braking level, steering wheel angle, cutting mechanism displacement, cutting mechanism rotation speed;

[0010] Using the continuous mining data to train an initial deep neural network model to obtain a trained deep neural network model;

[0011] S200, pre-drilling fracturing operation and rock formation strength parameter prediction: drilling test holes and vibrating the rock formation working surface through pre-drilling fracturing equipment, and inputting the pre-drilling technical parameters in this pre-drilling fracturing operation into the trained deep neural network model to calculate the rock formation strength parameters after the completion of this pre-drilling fracturing operation;

[0012] S300, determine a reasonable cutting depth control value: Considering the most unfavorable working condition of the pick of the continuous mining equipment and the durability of the pick material, the peak value of the repeated stress on the pick during operation should not be greater than the fatigue strength of the pick material, determine a reasonable cutting depth control value [ d t ], cutting depth control value[ d t The calculation of ] satisfies the following expression:

[0013]

[0014] in, σ rt is the average tensile strength of the rock formation after fracture, [ d t ] is the cutting depth control value, α is the half angle of the cutting tooth tip, f is the friction angle between the pick and the rock formation, S is the cross-sectional area at the most unfavorable position of the pick, [ σ sw ] is the fatigue strength limit value of the cutting tooth material; the cutting depth control value is used to specify the range of the cutting depth parameter, and the cutting depth parameter is the cutting depth value obtained by actual measurement;

[0015] S400, monitoring and control of continuous mining operations: During continuous mining operations, the pre-drilling technical parameters at different times, the rock formation strength parameters after pre-drilling fracturing, the mining environment status parameters, the continuous mining equipment status parameters and the cutting depth parameters are monitored in real time and input into the trained deep neural network model, and the continuous mining equipment operation strategies at different times are calculated to achieve control of the continuous mining operation parameters including the cutting depth.

[0016] Preferably, the training of the initial deep neural network model in step S100 specifically includes the following steps:

[0017] S110, training of rock formation strength parameters: after inputting the pre-drilling technical parameters in the continuous excavation data into the initial deep neural network model, calculating and outputting the rock formation strength parameters at the current moment, measuring the error between the predicted result and the true value according to the first loss function, and then guiding the learning process of the model until the first loss function meets the threshold;

[0018] S120, analogy of continuous mining equipment operation strategy: after the mining environment state parameters, the continuous mining equipment state parameters, the cutting depth parameters and the rock formation strength parameters in the continuous mining data are input into the initial deep neural network model, the continuous mining equipment control amount at the current moment is calculated and output; then the mining environment state parameters, the continuous mining equipment state parameters and the cutting depth parameters at the next moment are calculated according to the continuous mining equipment control amount at the current moment, and then the continuous mining equipment control amount at the next moment is calculated; and so on and so forth until the mining of the current rock formation working face is completed, at which time a series of the continuous mining equipment control amounts can be obtained, and this series of the continuous mining equipment control amounts is the continuous mining equipment operation strategy at the current moment;

[0019] S130. Training of continuous mining equipment operation strategy: calculating a second loss function evaluation value according to the continuous mining equipment operation strategy at the current moment; the second loss function evaluation value includes but is not limited to a difference evaluation value between the cutting depth at the current moment and the cutting depth control value, an evaluation value of the cutting ratio energy consumption at the current moment, an evaluation value of the cutting mechanism power at the current moment and an evaluation value of the flatness of the mining working surface; performing parameter optimization and updating of the deep neural network model at the current moment according to each of the second loss function evaluation values, guiding the learning process of the model until the second loss function evaluation value meets the threshold.

[0020] Preferably, in step S300, in order to ensure the flatness of the working surface after the mining operation, the cutting depth control value [ d t ]The following expression must be satisfied:

[0021] [ d t ]≥ h d

[0022] in, h d is the depth of the test hole.

[0023] Preferably, the monitoring and control of the continuous mining operation in step S400 further includes the following steps:

[0024] S410, measurement and control of cutting mechanism power: when it is learned through real-time monitoring that the cutting mechanism power exceeds the corresponding threshold, the control amount of the continuous mining equipment is controlled so that the continuous mining equipment stops the continuous mining operation; then the weight of the cutting mechanism power evaluation value at the current moment is increased, and according to steps S100 to S300, the deep neural network model is retrained, the rock formation strength parameters are predicted, and a reasonable cutting depth control value is determined, and then the continuous mining operation and its monitoring and control are carried out;

[0025] S420, measurement and control of the flatness of the mining working face: after completing one operation cycle of the rock working face mining operation, when real-time monitoring determines that the flatness of the mining working face exceeds the corresponding threshold, the control amount of the continuous mining equipment is controlled so that the continuous mining equipment stops the continuous mining operation; then the weight of the mining working face flatness evaluation value is increased, and according to steps S100 to S300, the deep neural network model is re-trained, the rock formation strength parameters are predicted, and a reasonable cutting depth control value is determined, and then continuous mining operations and their monitoring and control are carried out.

[0026] Preferably, the monitoring and control of the continuous mining operation in step S400 is implemented by a measurement and control system, which includes an information acquisition module, an information transmission module, a data storage module, an information processing module and a feedback module;

[0027] The information acquisition module includes a pre-drilling technical parameter acquisition unit, a mining environment state parameter acquisition unit, a continuous mining equipment state parameter acquisition unit and a cutting depth parameter acquisition unit, which are used to obtain monitoring data during the continuous mining operation through sensors installed on the continuous mining equipment;

[0028] The information transmission module is used to establish data connection and transmission between modules;

[0029] The data storage module is used to store historical data, data acquired by the information acquisition module, and data output by the trained deep neural network model;

[0030] The information processing module includes a data preprocessing unit, a deep neural network model unit and a judgment unit; wherein the data preprocessing unit is used to preprocess the original data stored in the data storage module to reduce data noise and redundancy; the deep neural network model unit is used to implement the function of the deep neural network model; the judgment unit is used to judge whether the data meets the threshold requirement;

[0031] The feedback module includes an early warning unit and a control unit; wherein the early warning unit is used to control the alarm to issue an early warning according to the judgment result of the information processing module; and the control unit is used to control the continuous acquisition device according to the judgment result of the information processing module.

[0032] Compared with the prior art, the beneficial effects of the present invention are as follows: for the automatic control of cutting depth during open-pit mining operations in hard rock environments, a method for automatic cutting depth control for continuous mining of hard rock formations based on deep learning is disclosed, the control method comprising the establishment and training of a deep neural network model, the prediction of pre-drilling fracturing operations and rock formation strength parameters, the determination of a reasonable cutting depth control value, and the monitoring and control of continuous mining operations; wherein, while the cuttability of hard rock formations is improved by pre-drilling fracturing operations, the rock formation strength parameters after the pre-drilling fracturing operations are predicted according to a deep learning algorithm, providing a quantitative indicator for the fracturing effect of pre-drilling fracturing operations on hard rock formations; and then a reasonable cutting depth value is determined according to the rock formation strength parameters after pre-drilling fracturing. It is beneficial to improve the mining efficiency of continuous mining equipment in hard rock environment and reduce the consumption of cutting tools; then, the current continuous mining equipment operation strategy is calculated in real time according to the rock strength parameters after pre-drilling and fracturing, real-time monitoring data and the trained deep neural network model, and then the real-time and continuous adjustment of the continuous mining equipment control quantity is carried out to realize real-time control of the cutting depth, which is beneficial to improve the control accuracy, safety and automation of the continuous mining equipment; finally, the cutting mechanism power and the flatness of the mining working surface are monitored according to the real-time monitoring data, and the weight of the deep neural network model for the cutting mechanism power or the flatness of the mining working surface is increased according to the real-time monitoring data to realize indirect control, so as to avoid overload operation of continuous mining equipment or failure to meet the quality of mining operations. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 This is a flow chart of the automatic cutting depth control method for continuous mining of hard rock strata based on deep learning of the present invention;

[0034] Figure 2 A schematic diagram of the structure of a deep neural network model provided by an embodiment of the present invention;

[0035] Figure 3 A schematic diagram of the structure of a continuous mining device provided by an embodiment of the present invention;

[0036] Figure 4 A schematic diagram of a cutting tooth of a cutting mechanism provided in one embodiment of the present invention;

[0037] Figure 5 A connection diagram of a measurement and control system provided in one embodiment of the present invention;

[0038] Figure markings: 11-cutting mechanism, 12-working face, 13-pre-drilling, 14-cutting teeth, 2-information acquisition module, 21-pre-drilling technical parameter acquisition unit, 22-mining environment state parameter acquisition unit, 23-continuous mining equipment state parameter acquisition unit, 24-cutting depth parameter acquisition unit, 3-information transmission module, 4-data storage module, 5-information processing module, 51-data pre-processing unit, 52-deep neural network model unit, 53-judgment unit, 6-feedback module, 61-early warning unit, 62-control unit. DETAILED DESCRIPTION

[0039] The following is combined with Figure 1-5 The embodiments of the present invention are described in more detail with reference to the accompanying drawings so that a person skilled in the art can implement the present invention after reading the present description. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0040] This application discloses Figures 1 to 5 The automatic cutting depth control method for continuous mining of hard rock formations based on deep learning shown in the figure specifically includes the following steps:

[0041] S100, Establishment and training of deep neural network model: Establish the initial deep neural network model, including the input layer ( I 1 , I 2 , … I k ), hidden layer( h 1 [1] , h 2 [1] ,… h n [3] ) and the output layer ( O 1 ); the input layer is used to receive data, the hidden layer is used for data processing and feature extraction, and the output layer is used to generate prediction results;

[0042] The data received by the input layer can be expressed as ( P 1 , P 2 , … P i ), including but not limited to normalized pre-drilling technical parameters, mining environment state parameters and continuous mining equipment state parameters at different times; the prediction results generated by the output layer are rock formation strength parameters and continuous mining equipment control quantities;

[0043] Continuous mining data under hard rock conditions are obtained by screening historical data, and the continuous mining data include but are not limited to pre-drilling technical parameters, mining environment status parameters, continuous mining equipment status parameters, cutting depth parameters, rock formation strength parameters and continuous mining equipment control quantities at different times; the pre-drilling technical parameters include but are not limited to hole depth, hole diameter, drilling hole distribution spacing, drilling energy consumption, vibration fracturing times and original rock formation strength parameters obtained from geological survey data; the mining environment status parameters include but are not limited to the geometric shape of the rock formation working face; the continuous mining equipment status parameters include but are not limited to the continuous mining equipment position, cutting mechanism linear speed, cutting mechanism angular speed, cutting mechanism output power; the rock formation strength parameters include but are not limited to the average tensile strength of the rock formation after fracturing; the continuous mining equipment control quantities include but are not limited to the throttle control opening, braking level, steering wheel angle, cutting mechanism displacement, cutting mechanism rotation speed;

[0044] The continuous mining data is used to train the initial deep neural network model to obtain a trained deep neural network model; the training of the initial deep neural network model specifically includes the following steps:

[0045] S110, training of rock formation strength parameters: after inputting the pre-drilling technical parameters in the continuous excavation data into the initial deep neural network model, calculating and outputting the rock formation strength parameters at the current moment, measuring the error between the predicted result and the true value according to the first loss function, and then guiding the learning process of the model until the first loss function meets the threshold;

[0046] S120, analogy of continuous mining equipment operation strategy: after the mining environment state parameters, the continuous mining equipment state parameters, the cutting depth parameters and the rock formation strength parameters in the continuous mining data are input into the initial deep neural network model, the continuous mining equipment control amount at the current moment is calculated and output; then the mining environment state parameters, the continuous mining equipment state parameters and the cutting depth parameters at the next moment are calculated according to the continuous mining equipment control amount at the current moment, and then the continuous mining equipment control amount at the next moment is calculated; and so on and so forth until the mining of the current rock formation working face is completed, at which time a series of the continuous mining equipment control amounts can be obtained, and this series of the continuous mining equipment control amounts is the continuous mining equipment operation strategy at the current moment;

[0047] S130, training of continuous mining equipment operation strategy: calculating a second loss function evaluation value according to the continuous mining equipment operation strategy at the current moment; the second loss function evaluation value includes but is not limited to a difference evaluation value between the cutting depth at the current moment and the cutting depth control value, a cutting specific energy consumption evaluation value at the current moment, a cutting mechanism power evaluation value at the current moment, and an evaluation value of the flatness of the mining working surface; performing parameter optimization and updating of the deep neural network model at the current moment according to each of the second loss function evaluation values, guiding the learning process of the model until the second loss function evaluation value meets the threshold;

[0048] S200, pre-drilling fracturing operation and rock formation strength parameter prediction: drilling test holes and vibrating the rock formation working surface through pre-drilling fracturing equipment, and inputting the pre-drilling technical parameters in this pre-drilling fracturing operation into the trained deep neural network model to calculate the rock formation strength parameters after the completion of this pre-drilling fracturing operation;

[0049] S300, determine a reasonable cutting depth control value: Considering the most unfavorable working condition of the pick of the continuous mining equipment and the durability of the pick material, the peak value of the repeated stress on the pick during operation should not be greater than the fatigue strength of the pick material, determine a reasonable cutting depth control value [ d t ], cutting depth control value[ d t The calculation of ] satisfies the following expression:

[0050] (1)

[0051] in, σ rt is the average tensile strength of the rock formation after fracture, [ d t ] is the cutting depth control value, α The half angle of the pick tip. f is the friction angle between the pick and the rock formation, S is the cross-sectional area at the most unfavorable position of the pick, [ σ sw ] is the fatigue strength limit value of the cutting tooth material; the cutting depth control value is used to specify the range of the cutting depth parameter, and the cutting depth parameter is the cutting depth value obtained by actual measurement;

[0052] In addition, in order to ensure the flatness of the working surface after mining, the cutting depth control value [ d t ]The following expression must be satisfied:

[0053] [ d t ]≥ hd (2)

[0054] in, h d is the depth of the test hole;

[0055] In the specific implementation, under typical working conditions, the average tensile strength of the rock formation after fracture σ rt 10MPa, half angle of pick tip α The friction angle between the pick and the rock formation is 17.5°. f The cross-sectional area at the most unfavorable position of the pick is 11.3°. S 706.5mm 2 , the fatigue strength limit value of the pick material [ σ sw ] is 500 MPa, and the cutting depth control value is [ d t ] According to formula (1), it is calculated as follows:

[0056] (3)

[0057] Test hole depth h d is 0.1m, combined with formula (2) and formula (3), the cutting depth control value [ d t ]The value range is 0.1m≤[ d t ]≤0.127m;

[0058] S400, Monitoring and control of continuous mining operations:

[0059] During the continuous mining operation, the pre-drilling technical parameters, the mining environment state parameters, the continuous mining equipment state parameters and the cutting depth parameters at different times are monitored in real time and input into the trained deep neural network model, the continuous mining equipment operation strategies at different times are calculated and used to control the continuous mining operation, so as to realize the control of the continuous mining operation parameters including the cutting depth; the monitoring and control of the continuous mining operation specifically includes the following steps:

[0060] S410, measurement and control of cutting mechanism power: when it is learned through real-time monitoring that the cutting mechanism power exceeds the corresponding threshold, the control amount of the continuous mining equipment is controlled so that the continuous mining equipment stops the continuous mining operation; then the weight of the cutting mechanism power evaluation value at the current moment is increased, and according to steps S100 to S300, the deep neural network model is retrained, the rock formation strength parameters are predicted, and a reasonable cutting depth control value is determined, and then the continuous mining operation and its monitoring and control are carried out;

[0061] S420, measurement and control of the flatness of the mining working face: after completing one operation cycle of the rock working face mining operation, when real-time monitoring determines that the flatness of the mining working face exceeds the corresponding threshold, the control amount of the continuous mining equipment is controlled so that the continuous mining equipment stops the continuous mining operation; then the weight of the mining working face flatness evaluation value is increased, and according to steps S100 to S300, the deep neural network model is re-trained, the rock formation strength parameters are predicted, and a reasonable cutting depth control value is determined, and then continuous mining operations and their monitoring and control are carried out.

[0062] In a specific implementation, the monitoring and control of the continuous mining operation in step S400 is implemented by a measurement and control system, which includes an information acquisition module 2, an information transmission module 3, a data storage module 4, an information processing module 5 and a feedback module 6;

[0063] The information acquisition module 2 includes a pre-drilling technical parameter acquisition unit 21, a mining environment state parameter acquisition unit 22, a continuous mining equipment state parameter acquisition unit 23 and a cutting depth parameter acquisition unit 24, which are used to obtain monitoring data during the continuous mining operation through sensors installed on the continuous mining equipment;

[0064] The information transmission module 3 is used to establish data connection and transmission between modules;

[0065] The data storage module 4 is used to store historical data, data acquired by the information acquisition module 2, and data output by the trained deep neural network model;

[0066] The information processing module 5 includes a data preprocessing unit 51, a deep neural network model unit 52 and a judgment unit 53; wherein the data preprocessing unit 51 is used to preprocess the original data stored in the data storage module 4 to reduce data noise and redundancy; the deep neural network model unit 52 is used to implement the function of the deep neural network model; the judgment unit 53 is used to judge whether the data meets the threshold requirement;

[0067] The feedback module 6 includes an early warning unit 61 and a control unit 62; wherein the early warning unit 61 is used to control the alarm to issue an early warning according to the judgment result of the information processing module 5; the control unit 62 is used to control the continuous collection device according to the judgment result of the information processing module 5.

[0068] It can be seen that the automatic cutting depth control method for continuous mining of hard rock strata based on deep learning can improve the cuttability of hard rock strata through pre-drilling and fracturing operations, and at the same time predict the rock strata strength parameters after the pre-drilling and fracturing operations are completed according to the deep learning algorithm, and determine the reasonable cutting depth value, thereby improving the mining efficiency of continuous mining equipment in hard rock environments and reducing the consumption of cutting tools. The current continuous mining equipment operation strategy is calculated in real time through the deep neural network model to achieve real-time control of the cutting depth, which is beneficial to improving the control accuracy, safety and automation of continuous mining equipment.

[0069] The above are one or more embodiments of the present invention, and the description is relatively specific and detailed, but it cannot be understood as limiting the scope of the patent of the present invention. It should be pointed out that for ordinary technicians in this field, several modifications and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be based on the attached claims.

Claims

1. A method for controlling the automatic cutting depth of continuous mining in hard rock formations based on deep learning, characterized in that: The following steps are involved: S100, establishment and training of deep neural network model: establish an initial deep neural network model, including input layer, hidden layer and output layer; The data received by the input layer include but are not limited to normalized pre-drilling technical parameters, mining environment state parameters and continuous mining equipment state parameters at different times; the prediction results generated by the output layer are rock formation strength parameters and continuous mining equipment control quantities; the rock formation strength parameters include but are not limited to the average tensile strength of the rock formation after fracturing; Filtering the historical data to obtain continuous mining data under hard rock conditions, the continuous mining data includes but is not limited to pre-drilling technical parameters, mining environment status parameters, continuous mining equipment status parameters, cutting depth parameters, rock formation strength parameters and continuous mining equipment control quantities at different times; Using the continuous mining data to train an initial deep neural network model to obtain a trained deep neural network model; S200, pre-drilling fracturing operation and rock formation strength parameter prediction: drilling test holes and vibrating the rock formation working surface through pre-drilling fracturing equipment, and inputting the pre-drilling technical parameters in this pre-drilling fracturing operation into the trained deep neural network model to calculate the rock formation strength parameters after the completion of this pre-drilling fracturing operation; S300, determine a reasonable cutting depth control value: cutting depth control value [ d t The calculation of ] satisfies the following expression: in, σ rt is the average tensile strength of the rock formation after fracture, [ d t ] is the cutting depth control value, α is the half angle of the cutting tooth tip, f is the friction angle between the pick and the rock formation, S is the cross-sectional area at the most unfavorable position of the pick, [ σ sw ] is the fatigue strength limit value of the cutting tooth material; the cutting depth control value is used to specify the range of the cutting depth parameter, and the cutting depth parameter is the cutting depth value obtained by actual measurement; S400, monitoring and control of continuous mining operations: During continuous mining operations, the pre-drilling technical parameters, the mining environment status parameters, the continuous mining equipment status parameters and the cutting depth parameters at different times are monitored in real time and input into the trained deep neural network model, and the continuous mining equipment operation strategies at different times are calculated and used to control the continuous mining operations.

2. The method for automatic cutting depth control in continuous excavation of hard rock strata based on deep learning according to claim 1 is characterized in that: The step S100 includes the following steps: S110, training of rock formation strength parameters: after inputting the pre-drilling technical parameters in the continuous excavation data into the initial deep neural network model, calculating and outputting the rock formation strength parameters at the current moment, measuring the error between the predicted result and the true value according to the first loss function, and then guiding the learning process of the model until the first loss function meets the threshold; S120, analogy of continuous mining equipment operation strategy: after the mining environment state parameters, the continuous mining equipment state parameters, the cutting depth parameters and the rock formation strength parameters in the continuous mining data are input into the initial deep neural network model, the continuous mining equipment control amount at the current moment is calculated and output; then the mining environment state parameters, the continuous mining equipment state parameters and the cutting depth parameters at the next moment are calculated according to the continuous mining equipment control amount at the current moment, and then the continuous mining equipment control amount at the next moment is calculated; and so on and so forth until the mining of the current rock formation working face is completed, at which time a series of the continuous mining equipment control amounts can be obtained, and this series of the continuous mining equipment control amounts is the continuous mining equipment operation strategy at the current moment; S130. Training of continuous mining equipment operation strategy: calculating a second loss function evaluation value according to the continuous mining equipment operation strategy at the current moment; the second loss function evaluation value includes but is not limited to a difference evaluation value between the cutting depth at the current moment and the cutting depth control value, an evaluation value of the cutting ratio energy consumption at the current moment, an evaluation value of the cutting mechanism power at the current moment and an evaluation value of the flatness of the mining working surface; performing parameter optimization and updating of the deep neural network model at the current moment according to each of the second loss function evaluation values, guiding the learning process of the model until the second loss function evaluation value meets the threshold.

3. The method for automatic cutting depth control in continuous mining of hard rock strata based on deep learning according to claim 1 is characterized in that: The cutting depth control value in step S300 is d t ]The following expression must be satisfied: [ d t ]≥ h d in, h d is the depth of the test hole.

4. The method for automatic cutting depth control in continuous mining of hard rock strata based on deep learning according to claim 1 is characterized in that: The step S400 includes the following steps: S410, measurement and control of cutting mechanism power: when it is learned through real-time monitoring that the cutting mechanism power exceeds the corresponding threshold, the control amount of the continuous mining equipment is controlled so that the continuous mining equipment stops the continuous mining operation; then the weight of the cutting mechanism power evaluation value at the current moment is increased, and according to steps S100 to S300, the deep neural network model is retrained, the rock formation strength parameters are predicted, and a reasonable cutting depth control value is determined, and then the continuous mining operation and its monitoring and control are carried out; S420, measurement and control of the flatness of the mining working face: after completing one operation cycle of the rock working face mining operation, when real-time monitoring determines that the flatness of the mining working face exceeds the corresponding threshold, the control amount of the continuous mining equipment is controlled so that the continuous mining equipment stops the continuous mining operation; then the weight of the mining working face flatness evaluation value is increased, and according to steps S100 to S300, the deep neural network model is re-trained, the rock formation strength parameters are predicted, and a reasonable cutting depth control value is determined, and then continuous mining operations and their monitoring and control are carried out.

5. The method for automatic cutting depth control in continuous mining of hard rock strata based on deep learning according to claim 1 is characterized in that: In step S100, the pre-drilling technical parameters include but are not limited to hole depth, hole diameter, drilling distribution spacing, drilling energy consumption, vibration cracking times and original rock strength parameters obtained from geological survey data; the mining environment state parameters include but are not limited to the geometric shape of the rock working face; the continuous mining equipment state parameters include but are not limited to the continuous mining equipment position, cutting mechanism linear speed, cutting mechanism angular velocity, cutting mechanism output power; the continuous mining equipment control quantity includes but is not limited to the throttle control opening, braking level, steering wheel angle, cutting mechanism displacement, and cutting mechanism rotation speed.

6. The method for automatic cutting depth control in continuous excavation of hard rock strata based on deep learning according to claim 1 is characterized in that: The monitoring and control of the continuous mining operation in step S400 is implemented by a measurement and control system, which includes an information acquisition module, an information transmission module, a data storage module, an information processing module and a feedback module; The information collection module includes a pre-drilling technical parameter collection unit, a mining environment state parameter collection unit, a continuous mining equipment state parameter collection unit and a cutting depth parameter collection unit; The information transmission module is used to establish data connection and transmission between modules; The data storage module is used to store historical data, data acquired by the information acquisition module, and data output by the trained deep neural network model; The information processing module includes a data preprocessing unit, a deep neural network model unit and a judgment unit; wherein the data preprocessing unit is used to preprocess the original data stored in the data storage module to reduce data noise and redundancy; the deep neural network model unit is used to implement the function of the deep neural network model; the judgment unit is used to judge whether the data meets the threshold requirement; The feedback module includes an early warning unit and a control unit; wherein the early warning unit is used to control the alarm to issue an early warning according to the judgment result of the information processing module; and the control unit is used to control the continuous acquisition device according to the judgment result of the information processing module.

Images