A control method of a cutter head in a TBM tunneling process

Abstract

The application relates to the field of tunnel excavation, and discloses a control method for a cutter head in a TBM (Tunnel Boring Machine) excavation process, which comprises the following steps: performing first excavation according to the operation parameters of the last shield, setting the operation parameters of the shield to preset operation parameters in a linear transition mode, then performing second excavation, and after the second excavation length reaches a threshold value or a preset time length, generating third excavation parameters based on the data collected by sensors in the second excavation process, linearly transitioning from the preset operation parameters to the third excavation parameters, and completing a shield stroke. The application can quickly build an auxiliary driving model in a new shield roadway. Through the above operation, the construction parameters can be quickly obtained in a poor geological environment, the safe operation range is obtained through the parameters, and the speed of the shield is improved.

Description

Technical Field

[0001] This invention relates to the field of tunnel boring, and more particularly to a method for controlling the cutterhead during TBM tunneling. Background Technology

[0002] Currently, the main methods for tunneling in coal mines include fully mechanized tunneling, drill-and-blast tunneling, and continuous miner tunneling (applicable only to coal mine tunneling). These methods are prone to imbalances in tunneling, anchoring, and hauling during actual construction.

[0003] In response to this situation, some coal mines have adopted tunnel boring machine (TBM) systems, which have greatly improved the efficiency of tunneling rock tunnels. The main equipment used in a TBM system is the tunnel boring machine (TBM), primarily used for tunnel excavation. Modern TBMs are technologically advanced, integrating multiple technologies such as optics, mechanics, electronics, hydraulics, and sensors. They can perform functions such as cutting and transporting soil and rock, and supporting the formed tunnel. They can be custom-designed to suit different geological conditions, resulting in very high overall reliability and safety.

[0004] However, tunnel boring machines (TBMs) require specialized personnel to operate, and TBMs used in coal mines necessitate personnel with considerable expertise. While intelligent assistance can effectively reduce the workload, the formation of coal mine roadways differs from that of conventional TBMs, necessitating specialized solutions. Coal mine roadway formation is primarily related to the geological environment, and there are few reports on optimizing operations based on this environment.

[0005] Due to the complexity of underground environments, the tunnel boring process in coal mine roadways must constantly pay attention to changes in the environment to avoid accidents caused by the use of models; for the operation of the cutterhead, geological differences should also be taken into account, but existing technologies rarely disclose this aspect. Summary of the Invention

[0006] The purpose of this invention is to overcome one or more of the above-mentioned existing technical problems and provide a method for controlling the cutterhead during TBM tunneling.

[0007] To achieve the above objectives, the present invention provides a method for controlling the cutterhead during TBM tunneling, comprising:

[0008] The first tunneling operation will proceed based on the last shield tunneling operation parameters;

[0009] The shield tunneling operation parameters are set to preset parameters using a linear transition method, and then the second tunneling is executed.

[0010] After the second tunneling length reaches the threshold or the preset time, the third tunneling parameters are generated based on the data collected by the sensors during the second tunneling process;

[0011] The tunnel boring machine (TBM) transitions linearly from the preset operating parameters to the third tunneling parameters, completing one tunneling cycle.

[0012] According to one aspect of the present invention, the third tunneling parameter is generated based on a long short-term memory neural network, wherein the input layer of the long short-term memory neural network includes collected vibration, sound and other data, and also includes a hidden layer and a fully connected layer. The hidden layer includes N storage units, and the fully connected layer includes a neuron with linear activation function. The target loss function is the Euclidean distance between the actual user-selected working condition and the predicted working condition.

[0013] According to one aspect of the invention, the length of the second tunneling is not less than 0.1m or the preset duration of the second tunneling is not less than 5s.

[0014] According to one aspect of the present invention, a dataset is generated based on historical shield tunneling data of a tunnel or similar tunnels, a long short-term memory neural network is trained based on the dataset, sensor data collected during the second tunneling process within 48 hours is obtained, and the sensor measurement value with the highest probability of occurrence is selected as the output of the long short-term memory neural network as a preset operating parameter.

[0015] According to one aspect of the present invention, during the formation of the tunnel, when the elevation change of the tunnel exceeds a height threshold, or when the tunnel boring machine changes from a stable section to an ascending or descending section, the manual operation parameters are re-collected, and preset operation parameters are set based on the manual operation parameters.

[0016] According to one aspect of the present invention, commonly used operating parameters for each category are determined according to the category of operating parameters, the probability of an operating parameter is determined according to the number of times it appears in the corresponding category, and the parameter with the highest occurrence is selected as the preset operating parameter.

[0017] According to one aspect of the present invention, commonly used operating parameters for each category are determined according to the category of operating parameters, the probability of an operating parameter is determined according to the number of times it appears in the corresponding category, and the parameter with the highest occurrence is selected as a candidate operating parameter.

[0018] The correlation between the various operation parameters is analyzed, and the operation parameter with the highest overall correlation is selected from the candidate operation parameters as the preset operation parameter.

[0019] According to one aspect of the invention, a fourth tunneling parameter is periodically generated based on data collected by sensors during the third tunneling process, and when the offset between the fourth tunneling parameter and the third tunneling parameter exceeds 20%, the system switches to manual driving mode.

[0020] According to one aspect of the invention, in response to a manual intervention command during the third tunneling process, the system switches to manual driving mode; and in manual driving mode, the parameters for manual driving are initialized to preset operating parameters.

[0021] To achieve the above objectives, the present invention provides a control system for the cutterhead during TBM tunneling, comprising:

[0022] First tunneling module: Performs the first tunneling based on the operating parameters of the last tunnel boring machine;

[0023] Second tunneling module: The shield tunneling operation parameters are set to preset operation parameters through a linear transition, and then the second tunneling is executed;

[0024] The third tunneling module: After the second tunneling length reaches the threshold or the preset time, it generates the third tunneling parameters based on the data collected by the sensors during the second tunneling process;

[0025] The tunnel boring machine (TBM) transitions linearly from the preset operating parameters to the third tunneling parameters, completing one tunneling cycle.

[0026] Based on this, the beneficial effects of the present invention are as follows: This application can realize the rapid construction of an assisted driving model in the tunnel of a new shield tunnel. Through the above operations, the tunnel parameters can be quickly obtained in a relatively poor geological environment. The safe operating range can be obtained through the parameters, thereby improving the speed of the shield tunnel. By setting the target loss function as the Euclidean distance between the actual user's selected working condition and the predicted working condition, the optimal selection of the user in the existing model library is realized. That is, through iteration, the consistency between the user's selection and the model's predicted value is achieved. Attached Figure Description

[0027] Figure 1 This is a flowchart of a cutterhead control method during TBM tunneling according to the present invention;

[0028] Figure 2 This is a flowchart of a cutterhead control system for TBM tunneling, according to the present invention. Detailed Implementation

[0029] The invention will now be discussed with reference to exemplary embodiments. It should be understood that the described embodiments are merely intended to enable those skilled in the art to better understand and thus implement the invention, and are not intended to imply any limitation on the scope of the invention.

[0030] As used herein, the term “comprising” and its variations are to be interpreted as open-ended terms meaning “including but not limited to”. The term “based on” is to be interpreted as “at least partially based on”, and the terms “one embodiment” and “an embodiment” are to be interpreted as “at least one embodiment”.

[0031] According to one embodiment of the present invention, Figure 1 This is a flowchart of a cutterhead control method during TBM tunneling, as described in this invention. Figure 1As shown, a method for controlling the cutterhead during TBM tunneling includes:

[0032] To achieve the above objectives, the present invention provides a method for controlling the cutterhead during TBM tunneling, comprising:

[0033] The first tunneling operation will proceed based on the last shield tunneling operation parameters;

[0034] The shield tunneling operation parameters are set to preset parameters using a linear transition method, and then the second tunneling is executed.

[0035] After the second tunneling length reaches the threshold or the preset time, the third tunneling parameters are generated based on the data collected by the sensors during the second tunneling process;

[0036] The tunnel boring machine (TBM) transitions linearly from the preset operating parameters to the third tunneling parameters, completing one tunneling cycle.

[0037] According to one embodiment of the present invention, the third tunneling parameter is generated based on a long short-term memory neural network, wherein the input layer of the long short-term memory neural network includes collected vibration, sound and other data, and also includes a hidden layer and a fully connected layer. The hidden layer includes N storage units, and the fully connected layer includes a neuron with linear activation function. The target loss function is the Euclidean distance between the actual user-selected working condition and the predicted working condition.

[0038] According to one embodiment of the present invention, the length of the second tunneling is not less than 0.1m or the preset duration of the second tunneling is not less than 5s.

[0039] According to one embodiment of the present invention, a dataset is generated based on historical shield tunneling data of the tunnel or similar tunnels. A long short-term memory neural network is trained based on the dataset to obtain sensor data collected during the second tunneling process within 48 hours. The sensor measurement value with the highest probability of occurrence is selected as the output of the long short-term memory neural network as a preset operating parameter.

[0040] According to one embodiment of the present invention, during the formation of the tunnel, when the elevation change of the tunnel exceeds the height threshold, or when the tunnel boring machine changes from the stable section to the ascending or descending section, the manual operation parameters are re-collected, and the preset operation parameters are set based on the manual operation parameters.

[0041] According to one embodiment of the present invention, commonly used operating parameters for each category are determined according to the category of operating parameters, the probability of an operating parameter is determined according to the number of times it appears in the corresponding category, and the parameter with the highest occurrence is selected as the preset operating parameter.

[0042] According to one embodiment of the present invention, commonly used operating parameters for each category are determined according to the category of operating parameters, the probability of an operating parameter is determined according to the number of times it appears in the corresponding category, and the parameter with the highest occurrence is selected as a candidate operating parameter.

[0043] The correlation between the various operation parameters is analyzed, and the operation parameter with the highest overall correlation is selected from the candidate operation parameters as the preset operation parameter.

[0044] According to one embodiment of the present invention, a fourth tunneling parameter is generated periodically based on data collected by sensors during the third tunneling process. When the offset between the fourth tunneling parameter and the third tunneling parameter exceeds 20%, the system switches to manual driving mode.

[0045] According to one embodiment of the present invention, in response to a manual intervention command during the third tunneling process, the system switches to manual driving mode; and in manual driving mode, the parameters for manual driving are initialized to preset operating parameters.

[0046] According to one embodiment of the present invention, the concept of the invention is that the working conditions of the previous operation are reasonable and acceptable, and the tunnel boring machine can use the working conditions to tunnel. It is further assumed that for the same type of coal and rock, if the same tunneling parameters are used, then through reasonable measurement, they should have consistent performance. Furthermore, by selecting a standard working condition, the feedback of the tunneling process under that working condition is determined, and based on this feedback, a suitable tunneling strategy and method are determined. After selecting a suitable tunneling strategy and method, the machine can gradually transition to the set parameters.

[0047] According to one embodiment of the present invention, sensors installed on the tunnel boring machine (TBM) are used to track the tunneling process. These sensors can be acoustic sensors to acquire acoustic signals, which can be obtained from various parts of the TBM, such as the cutterhead, shield, and support points in some embodiments of the invention. Acquiring these acoustic signals allows for the acquisition of energy absorption at the interface during tunneling, via vibration or sound waves, reflecting the contact between the cutterhead and the interface. Analyzing this energy absorption allows for a certain degree of reconstruction of the interface's properties. Filtering the acoustic signals allows for the acquisition of the acoustic distribution characteristics during tunneling. This filtering is performed to obtain the intensity of the sound field within a certain frequency range. It should be noted that the field strength of some acoustic sensors should be recalculated as the distance changes during tunneling. Furthermore, the representativeness of the sampling points should be carefully considered, as the noise level underground during TBM tunneling may be significant, and the sound distribution at some sampling points may be irrelevant to the tunneling process.

[0048] According to one embodiment of the present invention, by setting a minimum shield length for the tunnel boring machine, the problem of excessive drift between recommended and actual operations and frequent disconnection of assisted driving is avoided when the amount of data is too small.

[0049] According to one embodiment of the present invention, a training dataset is constructed by using tunnels with other approximate properties as a reference. When constructing the training dataset, a long short-term memory neural network is used to always prioritize the selection of operating parameters under new working conditions.

[0050] According to one embodiment of the present invention, the main parameters considered include altitude and interface type. The automatic driving model is stopped by determining changes in altitude. Significant changes in altitude correspond to changes in the tunnel boring machine's attitude, often involving soft coal at the shield interface. In such cases, the system should switch to manual driving mode. If necessary, industry experts can even be brought in to determine the corresponding operating parameters.

[0051] According to one embodiment of the present invention, since an operator may select multiple operations based on personal experience under similar working conditions, they should be prioritized and appropriate selections should be made.

[0052] According to one embodiment of the present invention, the model is constructed as follows: User operating parameters in response to standard operating conditions are collected; features are extracted from the user's operating parameters to reduce the dimensionality of the data, for example, by using convolution, or by mapping the user's operating parameters to a range and using that range as the operating feature; an operating feature sequence is constructed, wherein the value A of the operating feature sequence A is... i When the value is 1, it indicates that the user has selected this operational feature parameter. Using sensor data acquired under standard operating conditions as input and the operational feature sequence as output, an LSTM network is constructed and trained. The objective function used is to minimize the difference between the user's selection and the output value. The output of the trained network represents the probability of selecting a particular type of operational parameter. When determining a type of operational parameter, the actual operational parameter can be determined from its corresponding operational parameters. Approximately, this can be achieved by identifying commonly used operational parameters for each category, determining the probability of each operational parameter based on its frequency of occurrence in its corresponding category, and selecting the parameter with the highest frequency as a candidate operational parameter. This method avoids human influence.

[0053] According to one embodiment of the present invention, the algorithm for determining the correlation degree can be performed using a sensitivity algorithm. For example, the cutter head torque can be selected as the basis to determine the closest operating parameter A; then, the operation is repeated with any element in A to determine other operating parameters besides the cutter head torque and the set of operating parameters A.

[0054] According to one embodiment of the present invention, when only the third operating parameter is used, there may be situations in the middle that are far from the operator's intended operating value. In this case, by periodically determining the fourth operating parameter, although there are differences between the input sequence and the preset operating parameters in this process, abnormal operating conditions can be detected by setting a large deviation space.

[0055] Furthermore, to achieve the aforementioned objectives, this invention also provides a control system for the cutterhead during TBM tunneling. Figure 2 This is a flowchart of a cutterhead control system during TBM tunneling, as described in this invention. Figure 2 As shown, a control system for the cutterhead during TBM tunneling in this invention includes:

[0056] First tunneling module: Performs the first tunneling based on the operating parameters of the last tunnel boring machine;

[0057] Second tunneling module: The shield tunneling operation parameters are set to preset operation parameters through a linear transition, and then the second tunneling is executed;

[0058] The third tunneling module: After the second tunneling length reaches the threshold or the preset time, it generates the third tunneling parameters based on the data collected by the sensors during the second tunneling process;

[0059] The tunnel boring machine (TBM) transitions linearly from the preset operating parameters to the third tunneling parameters, completing one tunneling cycle.

[0060] According to one embodiment of the present invention, the third tunneling parameter is generated based on a long short-term memory neural network, wherein the input layer of the long short-term memory neural network includes collected vibration, sound and other data, and also includes a hidden layer and a fully connected layer. The hidden layer includes N storage units, and the fully connected layer includes a neuron with linear activation function. The target loss function is the Euclidean distance between the actual user-selected working condition and the predicted working condition.

[0061] According to one embodiment of the present invention, the length of the second tunneling is not less than 0.1m or the preset duration of the second tunneling is not less than 5s.

[0062] According to one embodiment of the present invention, a dataset is generated based on historical shield tunneling data of the tunnel or similar tunnels. A long short-term memory neural network is trained based on the dataset to obtain sensor data collected during the second tunneling process within 48 hours. The sensor measurement value with the highest probability of occurrence is selected as the output of the long short-term memory neural network as a preset operating parameter.

[0063] According to one embodiment of the present invention, during the formation of the tunnel, when the elevation change of the tunnel exceeds the height threshold, or when the tunnel boring machine changes from the stable section to the ascending or descending section, the manual operation parameters are re-collected, and the preset operation parameters are set based on the manual operation parameters.

[0064] According to one embodiment of the present invention, commonly used operating parameters for each category are determined according to the category of operating parameters, the probability of an operating parameter is determined according to the number of times it appears in the corresponding category, and the parameter with the highest occurrence is selected as the preset operating parameter.

[0065] According to one embodiment of the present invention, commonly used operating parameters for each category are determined according to the category of operating parameters, the probability of an operating parameter is determined according to the number of times it appears in the corresponding category, and the parameter with the highest occurrence is selected as a candidate operating parameter.

[0066] The correlation between the various operation parameters is analyzed, and the operation parameter with the highest overall correlation is selected from the candidate operation parameters as the preset operation parameter.

[0067] According to one embodiment of the present invention, a fourth tunneling parameter is generated periodically based on data collected by sensors during the third tunneling process. When the offset between the fourth tunneling parameter and the third tunneling parameter exceeds 20%, the system switches to manual driving mode.

[0068] According to one embodiment of the present invention, in response to a manual intervention command during the third tunneling process, the system switches to manual driving mode; and in manual driving mode, the parameters for manual driving are initialized to preset operating parameters.

[0069] According to one embodiment of the present invention, the concept of the invention is that the working conditions of the previous operation are reasonable and acceptable, and the tunnel boring machine can use the working conditions to tunnel. It is further assumed that for the same type of coal and rock, if the same tunneling parameters are used, then through reasonable measurement, they should have consistent performance. Furthermore, by selecting a standard working condition, the feedback of the tunneling process under that working condition is determined, and based on this feedback, a suitable tunneling strategy and method are determined. After selecting a suitable tunneling strategy and method, the machine can gradually transition to the set parameters.

[0070] According to one embodiment of the present invention, sensors installed on the tunnel boring machine (TBM) are used to track the tunneling process. These sensors can be acoustic sensors to acquire acoustic signals, which can be obtained from various parts of the TBM, such as the cutterhead, shield, and support points in some embodiments of the invention. Acquiring these acoustic signals allows for the acquisition of energy absorption at the interface during tunneling, via vibration or sound waves, reflecting the contact between the cutterhead and the interface. Analyzing this energy absorption allows for a certain degree of reconstruction of the interface's properties. Filtering the acoustic signals allows for the acquisition of the acoustic distribution characteristics during tunneling. This filtering is performed to obtain the intensity of the sound field within a certain frequency range. It should be noted that the field strength of some acoustic sensors should be recalculated as the distance changes during tunneling. Furthermore, the representativeness of the sampling points should be carefully considered, as the noise level underground during TBM tunneling may be significant, and the sound distribution at some sampling points may be irrelevant to the tunneling process.

[0071] According to one embodiment of the present invention, by setting a minimum shield length for the tunnel boring machine, the problem of excessive drift between recommended and actual operations and frequent disconnection of assisted driving is avoided when the amount of data is too small.

[0072] According to one embodiment of the present invention, a training dataset is constructed by using tunnels with other approximate properties as a reference. When constructing the training dataset, a long short-term memory neural network is used to always prioritize the selection of operating parameters under new working conditions.

[0073] According to one embodiment of the present invention, the main parameters considered include altitude and interface type. The automatic driving model is stopped by determining changes in altitude. Significant changes in altitude correspond to changes in the tunnel boring machine's attitude, often involving soft coal at the shield interface. In such cases, the system should switch to manual driving mode. If necessary, industry experts can even be brought in to determine the corresponding operating parameters.

[0074] According to one embodiment of the present invention, since an operator may select multiple operations based on personal experience under similar working conditions, they should be prioritized and appropriate selections should be made.

[0075] According to one embodiment of the present invention, the model is constructed as follows: User operation parameters in response to standard operating conditions are collected; features are extracted from these user operation parameters to reduce the dimensionality of the data, for example, using convolutional methods or mapping the user operation parameters to a range, which is then used as the operation feature; an operation feature sequence is constructed, where a value Ai of operation feature sequence A equals 1, indicating that the user has selected that operation feature parameter; sensor data acquired under standard operating conditions is used as input, and the operation feature sequence is used as output; an LSTM network is constructed and trained, with the objective function being to minimize the difference between the user's selection and the output value; the output of the trained network represents the probability of selecting a class of operation parameters. When determining a class of operation parameters, the actual operation parameter can be determined from its corresponding operation parameter. Approximately, commonly used operation parameters for each class are determined according to their categories, and their probabilities are determined according to the frequency of occurrence of each operation parameter in its corresponding category; the parameter with the highest occurrence is selected as the candidate operation parameter. This method avoids human influence.

[0076] According to one embodiment of the present invention, the algorithm for determining the correlation degree can be performed using a sensitivity algorithm. For example, the cutter head torque can be selected as the basis to determine the closest operating parameter A; then, the operation is repeated with any element in A to determine other operating parameters besides the cutter head torque and the set of operating parameters A.

[0077] According to one embodiment of the present invention, when only the third operating parameter is used, there may be situations in the middle that are far from the operator's intended operating value. In this case, by periodically determining the fourth operating parameter, although there are differences between the input sequence and the preset operating parameters in this process, abnormal operating conditions can be detected by setting a large deviation space.

[0078] Based on this, the beneficial effects of the present invention are that it can quickly build an assisted driving model in the tunnel of a new shield tunnel. Through the above operations, the tunnel parameters can be quickly obtained in adverse geological environments. The safe operating range can be obtained through the parameters, thereby improving the speed of the shield tunnel. By setting the target loss function as the Euclidean distance between the actual user's selected working condition and the predicted working condition, the optimal selection of the user in the existing model library is achieved. That is, through iteration, the consistency between the user's selection and the model's predicted value is achieved.

[0079] Those skilled in the art will recognize that the modules and algorithm steps described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0080] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the above-described apparatus and equipment can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0081] In the embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or modules may be electrical, mechanical, or other forms.

[0082] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical modules; that is, they may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected to achieve the purpose of the embodiments of the present invention, depending on actual needs.

[0083] In addition, the functional modules in the embodiments of the present invention can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0084] If the aforementioned functions are implemented as software functional modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, essentially, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the energy-saving signal transmission / reception methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, ROM, RAM, magnetic disks, or optical disks.

[0085] The above description is merely a preferred embodiment of this application and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in this application is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the inventive concept. For example, technical solutions formed by substituting the above features with (but not limited to) technical features with similar functions disclosed in this application.

[0086] It should be understood that the sequence number of each step in the invention and embodiments of the present invention does not absolutely imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

Claims

1. A method for controlling the cutterhead during TBM tunneling, characterized in that, include: The first tunneling operation will proceed based on the last shield tunneling operation parameters; The shield tunneling operation parameters are set to preset parameters using a linear transition method, and then the second tunneling is executed. After the second tunneling length reaches the threshold or the preset time, the third tunneling parameters are generated based on the data collected by the sensors during the second tunneling process; The tunnel boring machine (TBM) transitions linearly from the preset operating parameters to the third tunneling parameters, completing one tunneling cycle. The third tunneling parameters are generated based on a long short-term memory neural network. The input layer of the long short-term memory neural network includes the collected vibration and sound data. The long short-term memory neural network also includes a hidden layer and a fully connected layer. The hidden layer includes N storage units, and the fully connected layer includes a neuron with linear activation function. The target loss function is the Euclidean distance between the actual user-selected working condition and the predicted working condition. A dataset is generated based on historical shield tunneling data of the tunnel or similar tunnels. The long short-term memory neural network is trained based on the dataset. Sensor data collected during the second tunneling process within 48 hours is obtained. The sensor measurement value with the highest probability of occurrence is selected as the output of the long short-term memory neural network as the preset operating parameter. During the formation of the tunnel, when the elevation change of the tunnel exceeds the height threshold, or when the tunnel boring machine changes from the stable section to the ascending or descending section, the manual operation parameters are collected again, and the preset operation parameters are set based on the manual operation parameters. The fourth tunneling parameter is generated periodically based on the data collected by the sensors during the third tunneling process. When the deviation between the fourth tunneling parameter and the third tunneling parameter exceeds 20%, the system switches to manual driving mode.

2. The method for controlling the cutterhead during TBM tunneling as described in claim 1, characterized in that, The length of the second tunneling shall not be less than 0.1m or the preset duration of the second tunneling shall not be less than 5s.

3. The method for controlling the cutterhead during TBM tunneling as described in claim 1, characterized in that, The commonly used operating parameters for each category are determined according to the category of operating parameters. The probability of an operating parameter is determined according to the number of times it appears in the corresponding category. The parameter with the highest occurrence is selected as the preset operating parameter.

4. The method for controlling the cutterhead during TBM tunneling as described in claim 3, characterized in that, The commonly used operation parameters for each category are determined according to the category of operation parameters. The probability of an operation parameter is determined according to the number of times it appears in the corresponding category. The parameter with the highest occurrence is selected as the candidate operation parameter. The correlation between each operation parameter is analyzed, and the operation parameter with the highest overall correlation is selected as the preset operation parameter from the candidate operation parameters.

5. The method for controlling the cutterhead during TBM tunneling as described in claim 1, characterized in that, In response to the instruction for manual intervention during the third tunneling process, the system switches to manual driving mode; and in manual driving mode, the parameters for manual driving are initialized to preset operating parameters.

6. A control system for the cutterhead during TBM tunneling, used to implement the method described in any one of claims 1 to 5, characterized in that, include: First tunneling module: Performs the first tunneling based on the operating parameters of the last tunnel boring machine; Second tunneling module: The shield tunneling operation parameters are set to preset operation parameters through a linear transition, and then the second tunneling is executed; The third tunneling module: After the second tunneling length reaches the threshold or the preset time, it generates the third tunneling parameters based on the data collected by the sensors during the second tunneling process; The tunnel boring machine (TBM) transitions linearly from the preset operating parameters to the third tunneling parameters, completing one tunneling cycle.

Images