A method for attitude monitoring during TBM tunneling
By acquiring real-time parameters and constructing an offset prediction model during the tunnel boring machine's excavation process, and combining acoustic signal and information entropy analysis, the problem of monitoring the attitude change of the tunnel boring machine was solved, enabling precise monitoring of the attitude and adjustment of the support strategy, thereby improving the safety and efficiency of the tunneling process.
Patent Information
- Application Number
- CN202311085475.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2023-08-25
- Publication Date
- 2026-11-17
- Estimated Expiration
- 2043-08-25
AI Technical Summary
Existing technologies cannot effectively monitor the attitude changes of tunnel boring machines during the excavation of coal mine roadways, especially in complex geological environments, particularly in soft coal seams and layered, severely fractured rock strata. Traditional support methods cannot provide attitude-related information.
By acquiring real-time parameters and shield machine operating parameters during tunnel excavation, supervised and unsupervised offset prediction models are constructed. Combined with acoustic signal and information entropy analysis, changes in tunnel type and attitude are monitored, and adjustments to the support area are predicted through machine learning and mechanical models.
It enables real-time monitoring of the tunnel boring machine's attitude changes, improves the safety and efficiency of the tunneling process, reduces mechanical control lag, and provides precise support strategies.
Smart Images

Figure CN117189147B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of tunnel boring, and more particularly to a method for attitude monitoring during TBM tunneling. Background Technology
[0002] Currently, tunnel boring machines are commonly used to form tunnels. In recent years, due to their fast excavation speed and high safety factor, tunnel boring machines have been used to form roadways or inclined shafts.
[0003] Because the formation of coal mine roadways involves changes in the geological environment, and during the tunneling process, complex environments such as soft coal seams are often encountered, and the rock strata where coal mine roadways are located are mostly layered with severe joints and fissures, especially in deep roadways.
[0004] In existing technologies, force-supported anchor bolts and anti-scour brackets are commonly used to support tunnels. However, these methods cannot directly provide attitude-related information. Such information, besides coming from sensors mounted on the tunnel boring machine (TBM), also needs to consider factors such as the characteristics of the tunnel. When using a TBM to form a tunnel, existing solutions cannot effectively address the attitude monitoring problem. Summary of the Invention
[0005] The purpose of this invention is to overcome one or more of the above-mentioned existing technical problems and provide a method for attitude monitoring during TBM tunneling.
[0006] To achieve the above objectives, the present invention provides an attitude monitoring method for TBM tunneling, comprising:
[0007] When the tunnel excavation length is shorter than the threshold, the real-time parameters of the tunnel and the working parameters of the tunnel boring machine are obtained.
[0008] When the tunnel excavation length first exceeds the threshold, a supervised offset prediction model is constructed. The input of the offset prediction model is the working parameters of the tunnel boring machine and the vibration of the cutterhead, and the output is the offset value of the cutterhead in the vertical direction.
[0009] Based on the working parameters and acoustic signals of the tunnel boring machine during the tunneling process, it is determined whether a change in the tunnel category has occurred. When a change in tunnel category has occurred, the data of the new category is labeled and retrained.
[0010] The trained offset prediction model is used to predict the offset of the cutterhead in the vertical direction. The information entropy during the tunneling process is calculated based on the predicted value, and the attitude is predicted based on the information entropy.
[0011] According to one aspect of the invention, the instantaneous parameters of the roadway include the pressure and displacement values of the roadway sidewalls, roof, and floor.
[0012] The operating parameters of a tunnel boring machine (TBM) include the cutterhead's rotational speed, torque, power, cutterhead propulsion power, and thrust during the tunneling process; the values of the horizontal gyroscopes installed on the shield and each functional trolley section; and the pressure values at the TBM-tunnel contact surface.
[0013] According to one aspect of the present invention, the working parameters of the tunnel boring machine are used as input to a supervised offset prediction model to obtain the working condition category corresponding to the tunnel boring machine excavation process;
[0014] Clustering is performed based on the intensity of the acquired acoustic signals. The cluster labels are the operating condition categories. The cluster in which the operating condition belongs is determined based on the currently acquired acoustic signals, and whether it is consistent with the operating condition category is determined based on the distance from the cluster center.
[0015] According to one aspect of the invention, when the roadway category changes, manual operation is used instead of automatic operation, and the roadway category and working condition are marked after a tunneling process is completed. After the marking is completed, incremental training is performed.
[0016] According to one aspect of the present invention, when the tunnel excavation length first exceeds a threshold and the cumulative excavation length of the automated driving exceeds the automated driving threshold, an unsupervised offset prediction model is constructed. The input of the offset prediction model is the working parameters of the tunnel boring machine and the historical data of the cutterhead vibration. The output of the offset prediction model is the offset value of the tunnel boring machine cutterhead in space. Based on the output of the unsupervised offset prediction model, it is determined whether there is a need to switch to manual driving.
[0017] According to one aspect of the present invention, the response values of different parts of the tunnel to pressure and facilities are obtained based on the mechanical model of similar tunnels in the mining area, the displacement probability of multiple parts in the shield tunnel is determined based on the mechanical response values, and the information entropy is calculated based on the actual offset of each part in various events.
[0018] According to one aspect of the present invention, when determining attitude changes based on information entropy, the support area is determined according to the position corresponding to the attitude change, and the support strategy of the corresponding part is changed.
[0019] According to one aspect of the present invention, monitoring data of different types of tunnels after their formation are obtained, the deformation direction and magnitude of the tunnels in response to sound vibration and tunneling time after their formation are statistically analyzed, the tunnel deformation probability distribution associated with shield parameters is obtained, the information entropy is calculated based on the probability distribution of tunnels of the same type, and the consistency of the categories is determined based on the difference in information entropy.
[0020] According to one aspect of the present invention, the vibration of each zone corresponding to the actual working parameters of the cutterhead is acquired, and the vibration frequency, vibration acceleration and vibration velocity of the effective wave of the shield cutterhead vibration signal are acquired; the vibration signal of the cutterhead is collected according to the different angles of the cutterhead corresponding to the corresponding zones and the real-time working conditions, and normalized using the Softmax function.
[0021] To achieve the above objectives, the present invention provides an attitude monitoring system for TBM tunneling, comprising:
[0022] Real-time parameter acquisition module: When the tunnel excavation length is shorter than the threshold, it acquires the instantaneous parameters of the tunnel and the working parameters of the tunnel boring machine.
[0023] Offset prediction model building module: When the tunnel excavation length first exceeds the threshold, a supervised offset prediction model is built. The input of the offset prediction model is the working parameters of the tunnel boring machine and the vibration of the cutterhead, and the output is the offset value of the cutterhead in the vertical direction.
[0024] Tunnel Category Confirmation Module: Based on the working parameters and acoustic signals of the tunnel boring machine during the tunneling process, determine whether a change in tunnel category has occurred. When a tunnel category shift occurs, label the data of the new category and retrain.
[0025] Attitude change confirmation module: The trained offset prediction model is used to predict the offset of the cutterhead in the vertical direction, and the information entropy during the tunneling process is calculated based on the predicted value. Based on the information entropy, it is predicted whether the attitude has changed.
[0026] Based on this, the beneficial effect of the present invention is that it solves the problem that the attitude change of the tunnel boring machine cannot be known through traditional technology. Attached Figure Description
[0027] Figure 1 This is a flowchart of an attitude monitoring method for TBM tunneling process according to the present invention;
[0028] Figure 2 This is a flowchart of an attitude monitoring 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 an attitude monitoring method for a TBM tunneling process according to the present invention, as shown below. Figure 1 As shown, a method for attitude monitoring during TBM tunneling includes:
[0032] To achieve the above objectives, the present invention provides an attitude monitoring method for TBM tunneling, comprising:
[0033] When the tunnel excavation length is shorter than the threshold, the real-time parameters of the tunnel and the working parameters of the tunnel boring machine are obtained.
[0034] When the tunnel excavation length first exceeds the threshold, a supervised offset prediction model is constructed. The input of the offset prediction model is the working parameters of the tunnel boring machine and the vibration of the cutterhead, and the output is the offset value of the cutterhead in the vertical direction.
[0035] Based on the working parameters and acoustic signals of the tunnel boring machine during the tunneling process, it is determined whether a change in the tunnel category has occurred. When a change in tunnel category has occurred, the data of the new category is labeled and retrained.
[0036] The trained offset prediction model is used to predict the offset of the cutterhead in the vertical direction. The information entropy during the tunneling process is calculated based on the predicted value, and the attitude is predicted based on the information entropy.
[0037] According to one embodiment of the present invention, the instantaneous parameters of the roadway include the pressure and displacement values of the roadway sidewalls, roof, and floor.
[0038] The operating parameters of a tunnel boring machine (TBM) include the cutterhead's rotational speed, torque, power, cutterhead propulsion power, and thrust during the tunneling process; the values of the horizontal gyroscopes installed on the shield and each functional trolley section; and the pressure values at the TBM-tunnel contact surface.
[0039] According to one embodiment of the present invention, the working parameters of the tunnel boring machine are used as input to a supervised offset prediction model to obtain the working condition category corresponding to the tunnel boring machine excavation process;
[0040] Clustering is performed based on the intensity of the acquired acoustic signals. The cluster labels are the operating condition categories. The cluster in which the operating condition belongs is determined based on the currently acquired acoustic signals, and whether it is consistent with the operating condition category is determined based on the distance from the cluster center.
[0041] According to one embodiment of the present invention, when the roadway category changes, manual operation is used instead of automatic operation, and the roadway category and working condition are marked after a tunneling process is completed. After the marking is completed, incremental training is performed.
[0042] According to one embodiment of the present invention, when the tunnel excavation length first exceeds a threshold and the cumulative excavation length of the automated driving exceeds the automated driving threshold, an unsupervised offset prediction model is constructed. The input of the offset prediction model is the working parameters of the tunnel boring machine and the historical data of the cutterhead vibration. The output of the offset prediction model is the offset value of the tunnel boring machine cutterhead in space. Based on the output of the unsupervised offset prediction model, it is determined whether there is a need to switch to manual driving.
[0043] According to one embodiment of the present invention, the response values of different parts of the tunnel to pressure and facilities are obtained based on the mechanical model of similar tunnels in the mining area. The displacement probability of multiple parts in the shield tunnel is determined based on the mechanical response values. The information entropy is calculated based on the actual offset of each part in various events.
[0044] According to one embodiment of the present invention, when determining the attitude change based on information entropy, the support area is determined according to the position corresponding to the attitude change, and the support strategy of the corresponding part is changed.
[0045] According to one embodiment of the present invention, monitoring data of different types of tunnels after their formation are obtained, the deformation direction and magnitude of the tunnels in response to sound vibration and tunneling time after their formation are statistically analyzed, the tunnel deformation probability distribution associated with shield parameters is obtained, the information entropy is calculated based on the probability distribution of tunnels of the same type, and the consistency of the categories is determined based on the difference in information entropy.
[0046] According to one embodiment of the present invention, the vibration of each zone corresponding to the actual working parameters of the cutterhead is acquired, and the vibration frequency, vibration acceleration and vibration velocity of the effective wave of the shield cutterhead vibration signal are acquired; the vibration signal of the cutterhead is collected according to the different angles of the cutterhead corresponding to the corresponding zones and the real-time working conditions, and normalized using the Softmax function.
[0047] According to one embodiment of the present invention, when the tunnel excavation length is shorter than a threshold, since the number of datasets is insufficient for training, and generally the properties of the tunnel differ from those of other tunnels, autonomous driving is generally not carried out by transplanting data and models from other tunnels. Furthermore, even if the model is updated through incremental training, it may cause errors in some key operations because the model has already learned the characteristics of other tunnels.
[0048] According to one embodiment of the present invention, when the tunnel excavation length first exceeds a threshold, the size of the dataset is sufficient to meet the requirements. At this time, a supervised offset prediction model is constructed, wherein the input of the offset prediction model is the operating parameters of the tunnel boring machine. These operating parameters may include common parameters. Some fixed operating parameters that do not change with the geological environment should not be included in the training range. The output is the offset value of the cutterhead in the vertical direction. This value can be adjusted according to changes in the operating parameters and can also provide information on the location. The specific threshold length can be set as needed. For tunnels of new geological types, this length is generally greater than 60m.
[0049] According to one embodiment of the present invention, it is determined whether a change in roadway category has occurred based on the working parameters and acoustic signals of the tunnel boring machine during the tunneling process. This process is carried out through a supervised network model, which provides a type determined based on parameters and environmental parameters. This type is trained with prior data, which may have missing geological types, resulting in a shift in roadway category. When a shift in roadway category occurs, the data of the new category is labeled and retrained.
[0050] According to one embodiment of the present invention, a trained offset prediction model is used to predict the value of the offset of the cutterhead in the vertical direction to obtain a reference value of the possible position of the cutterhead in the shield direction under the condition of model matching. This value is used to calculate the information entropy in the shield tunneling process, that is, to predict the probability of the event by this prediction and to determine the potential information entropy based on the probability of the event.
[0051] According to one embodiment of the present invention, the information entropy during the tunneling process is calculated based on the predicted value. When the information entropy changes, it is possible to predict whether a change in attitude has occurred and whether there is a corresponding risk based on the information value.
[0052] According to one embodiment of the present invention, the collected real-time parameters of the tunnel are related to the impact of the tunnel boring machine (TBM) on the tunnel, and some properties of the tunnel also affect the state of the TBM. By collecting relevant parameters, such as tracking changes in displacement, attitude monitoring can be achieved. Similarly, the operating parameters of the TBM also affect the specific attitude. During the tunnel boring process, the operator can adjust the parameters of the TBM propulsion system and adjust the TBM attitude based on manual experience or a certain control strategy, according to the TBM posture deviation measured by the TBM guidance system. However, due to the large mass of the TBM, there is a certain lag in mechanical control. Nevertheless, the operating parameters set by the operator or the actual operating parameters set can be tracked within a certain period of time. Therefore, when the geological environment is basically stable or its composition does not change significantly within a certain range, machine learning can be used to make certain predictions about the impact of operating parameters on attitude.
[0053] According to one embodiment of the present invention, common supervised learning models can be random forest models, support vector machine (SVM) models, Xgboost models, LSTM models, etc.
[0054] According to one embodiment of the present invention, after acquiring sample data, the operator can remove erroneous data or accidental and unrepresentative sample data to obtain representative data. In cases where sensor data values are missing, fitting can be performed based on the sensor values at times adjacent to the missing time point. The above method may result in data loss, but the corresponding data can be placed into the test set, not the training set.
[0055] According to one embodiment of the present invention, an acoustic signal is formed based on the output of another working parameter of the tunnel, and clustering this signal can obtain the tunnel category. For each data block, a density-based spatial clustering (DBSCAN) algorithm is used for cluster learning and the clustering result is determined. Based on the results of geological exploration, the basic geological type can be determined, and once the geological type is determined, it can be used as the number of cluster centers.
[0056] According to one embodiment of the present invention, K-Means clustering is performed based on the energy distribution characteristics obtained by ultrasonic detection, and the clustering results are analyzed to obtain the clustering results. When the type of roadway changes, the existing data can no longer meet the needs of autonomous driving, thus requiring manual intervention to perform the operation. During manual intervention, the roadway excavation process can be tracked, and data can be selectively collected or all data can be collected to update the model.
[0057] According to one embodiment of the present invention, by referring to the information entropy value under other conditions, a predicted risk distribution can be obtained based on the predicted change in attitude. Compared with the situation where the model is lacking and an accurate information distribution cannot be obtained, the actual attitude distribution serves as the calculation benchmark for information entropy. Therefore, in the process of error accumulation, as long as the information entropy does not deteriorate, it can indicate that the attitude has not deteriorated.
[0058] According to one embodiment of the present invention, in some cases, such as through geological models, stress distribution over a certain period of time can be predicted; or based on finite element algorithms, risk distribution at interfaces can be obtained; or based on risk records of similar roadways in coal mines, especially data obtained from anchor bolt measurements within roadways, statistical analysis of these data can yield the probability distribution of events. For example, there is a correlation between structural spalling and pressure; the probability of spalling can be obtained by analyzing the trend of pressure distribution. Furthermore, the distribution of events such as spalling and the displacement changes of the roof and floor slabs are similar, both being parameters affecting the attitude of the tunnel boring machine. The present invention obtains the information entropy of risk by tracking this data.
[0059] According to one embodiment of the present invention, the information entropy can be calculated in sections based on the working functional range of the tunnel boring machine, or it can be calculated as a whole. When calculating based on sections, if the information entropy of a region changes significantly and indicates a corresponding risk, it can be addressed by strengthening the support. For example, in the front section, in addition to the front shield, it can also be addressed by adding support and netting.
[0060] According to one embodiment of the present invention, by monitoring the deformation direction and deformation magnitude of the formed tunnel, the probability distribution of tunnel deformation is determined. When the pressure is approximate, the probability distribution assumes that the corresponding events have a consistent probability and is used to track the changes in the attitude of the tunnel boring machine.
[0061] Furthermore, to achieve the aforementioned objectives, this invention also provides an attitude monitoring system for TBM tunneling. Figure 2 This is a flowchart of an attitude monitoring system during the TBM tunneling process according to the present invention, as shown below. Figure 2 As shown, an attitude monitoring system for TBM tunneling in this invention includes:
[0062] Real-time parameter acquisition module: When the tunnel excavation length is shorter than the threshold, it acquires the instantaneous parameters of the tunnel and the working parameters of the tunnel boring machine.
[0063] Offset prediction model building module: When the tunnel excavation length first exceeds the threshold, a supervised offset prediction model is built. The input of the offset prediction model is the working parameters of the tunnel boring machine and the vibration of the cutterhead, and the output is the offset value of the cutterhead in the vertical direction.
[0064] Tunnel Category Confirmation Module: Based on the working parameters and acoustic signals of the tunnel boring machine during the tunneling process, determine whether a change in tunnel category has occurred. When a tunnel category shift occurs, label the data of the new category and retrain.
[0065] Attitude change confirmation module: The trained offset prediction model is used to predict the offset of the cutterhead in the vertical direction, and the information entropy during the tunneling process is calculated based on the predicted value. Based on the information entropy, it is predicted whether the attitude has changed.
[0066] According to one embodiment of the present invention, the instantaneous parameters of the roadway include the pressure and displacement values of the roadway sidewalls, roof, and floor.
[0067] The operating parameters of a tunnel boring machine (TBM) include the cutterhead's rotational speed, torque, power, cutterhead propulsion power, and thrust during the tunneling process; the values of the horizontal gyroscopes installed on the shield and each functional trolley section; and the pressure values at the TBM-tunnel contact surface.
[0068] According to one embodiment of the present invention, the working parameters of the tunnel boring machine are used as input to a supervised offset prediction model to obtain the working condition category corresponding to the tunnel boring machine excavation process;
[0069] Clustering is performed based on the intensity of the acquired acoustic signals. The cluster labels are the operating condition categories. The cluster in which the operating condition belongs is determined based on the currently acquired acoustic signals, and whether it is consistent with the operating condition category is determined based on the distance from the cluster center.
[0070] According to one embodiment of the present invention, when the roadway category changes, manual operation is used instead of automatic operation, and the roadway category and working condition are marked after a tunneling process is completed. After the marking is completed, incremental training is performed.
[0071] According to one embodiment of the present invention, when the tunnel excavation length first exceeds a threshold and the cumulative excavation length of the automated driving exceeds the automated driving threshold, an unsupervised offset prediction model is constructed. The input of the offset prediction model is the working parameters of the tunnel boring machine and the historical data of the cutterhead vibration. The output of the offset prediction model is the offset value of the tunnel boring machine cutterhead in space. Based on the output of the unsupervised offset prediction model, it is determined whether there is a need to switch to manual driving.
[0072] According to one embodiment of the present invention, the response values of different parts of the tunnel to pressure and facilities are obtained based on the mechanical model of similar tunnels in the mining area. The displacement probability of multiple parts in the shield tunnel is determined based on the mechanical response values. The information entropy is calculated based on the actual offset of each part in various events.
[0073] According to one embodiment of the present invention, when determining the attitude change based on information entropy, the support area is determined according to the position corresponding to the attitude change, and the support strategy of the corresponding part is changed.
[0074] According to one embodiment of the present invention, monitoring data of different types of tunnels after their formation are obtained, the deformation direction and magnitude of the tunnels in response to sound vibration and tunneling time after their formation are statistically analyzed, the tunnel deformation probability distribution associated with shield parameters is obtained, the information entropy is calculated based on the probability distribution of tunnels of the same type, and the consistency of the categories is determined based on the difference in information entropy.
[0075] According to one embodiment of the present invention, the vibration of each zone corresponding to the actual working parameters of the cutterhead is acquired, and the vibration frequency, vibration acceleration and vibration velocity of the effective wave of the shield cutterhead vibration signal are acquired; the vibration signal of the cutterhead is collected according to the different angles of the cutterhead corresponding to the corresponding zones and the real-time working conditions, and normalized using the Softmax function.
[0076] According to one embodiment of the present invention, when the tunnel excavation length is shorter than a threshold, since the number of datasets is insufficient for training, and generally the properties of the tunnel differ from those of other tunnels, autonomous driving is generally not carried out by transplanting data and models from other tunnels. Furthermore, even if the model is updated through incremental training, it may cause errors in some key operations because the model has already learned the characteristics of other tunnels.
[0077] According to one embodiment of the present invention, when the tunnel excavation length first exceeds a threshold, the size of the dataset is sufficient to meet the requirements. At this time, a supervised offset prediction model is constructed, wherein the input of the offset prediction model is the operating parameters of the tunnel boring machine. These operating parameters may include common parameters. Some fixed operating parameters that do not change with the geological environment should not be included in the training range. The output is the offset value of the cutterhead in the vertical direction. This value can be adjusted according to changes in the operating parameters and can also provide information on the location. The specific threshold length can be set as needed. For tunnels of new geological types, this length is generally greater than 60m.
[0078] According to one embodiment of the present invention, it is determined whether a change in roadway category has occurred based on the working parameters and acoustic signals of the tunnel boring machine during the tunneling process. This process is carried out through a supervised network model, which provides a type determined based on parameters and environmental parameters. This type is trained with prior data, which may have missing geological types, resulting in a shift in roadway category. When a shift in roadway category occurs, the data of the new category is labeled and retrained.
[0079] According to one embodiment of the present invention, a trained offset prediction model is used to predict the value of the offset of the cutterhead in the vertical direction to obtain a reference value of the possible position of the cutterhead in the shield direction under the condition of model matching. This value is used to calculate the information entropy in the shield tunneling process, that is, to predict the probability of the event by this prediction and to determine the potential information entropy based on the probability of the event.
[0080] According to one embodiment of the present invention, the information entropy during the tunneling process is calculated based on the predicted value. When the information entropy changes, it is possible to predict whether a change in attitude has occurred and whether there is a corresponding risk based on the information value.
[0081] According to one embodiment of the present invention, the collected real-time parameters of the tunnel are related to the impact of the tunnel boring machine (TBM) on the tunnel, and some properties of the tunnel also affect the state of the TBM. By collecting relevant parameters, such as tracking changes in displacement, attitude monitoring can be achieved. Similarly, the operating parameters of the TBM also affect the specific attitude. During the tunnel boring process, the operator can adjust the parameters of the TBM propulsion system and adjust the TBM attitude based on manual experience or a certain control strategy, according to the TBM posture deviation measured by the TBM guidance system. However, due to the large mass of the TBM, there is a certain lag in mechanical control. Nevertheless, the operating parameters set by the operator or the actual operating parameters set can be tracked within a certain period of time. Therefore, when the geological environment is basically stable or its composition does not change significantly within a certain range, machine learning can be used to make certain predictions about the impact of operating parameters on attitude.
[0082] According to one embodiment of the present invention, common supervised learning models can be random forest models, support vector machine (SVM) models, Xgboost models, LSTM models, etc.
[0083] According to one embodiment of the present invention, after acquiring sample data, the operator can remove erroneous data or accidental and unrepresentative sample data to obtain representative data. In cases where sensor data values are missing, fitting can be performed based on the sensor values at times adjacent to the missing time point. The above method may result in data loss, but the corresponding data can be placed into the test set, not the training set.
[0084] According to one embodiment of the present invention, an acoustic signal is formed based on the output of another working parameter of the tunnel, and clustering this signal can obtain the tunnel category. For each data block, a density-based spatial clustering (DBSCAN) algorithm is used for cluster learning and the clustering result is determined. Based on the results of geological exploration, the basic geological type can be determined, and once the geological type is determined, it can be used as the number of cluster centers.
[0085] According to one embodiment of the present invention, K-Means clustering is performed based on the energy distribution characteristics obtained by ultrasonic detection, and the clustering results are analyzed to obtain the clustering results. When the type of roadway changes, the existing data can no longer meet the needs of autonomous driving, thus requiring manual intervention to perform the operation. During manual intervention, the roadway excavation process can be tracked, and data can be selectively collected or all data can be collected to update the model.
[0086] According to one embodiment of the present invention, by referring to the information entropy value under other conditions, a predicted risk distribution can be obtained based on the predicted change in attitude. Compared with the situation where the model is lacking and an accurate information distribution cannot be obtained, the actual attitude distribution serves as the calculation benchmark for information entropy. Therefore, in the process of error accumulation, as long as the information entropy does not deteriorate, it can indicate that the attitude has not deteriorated.
[0087] According to one embodiment of the present invention, in some cases, such as through geological models, stress distribution over a certain period of time can be predicted; or based on finite element algorithms, risk distribution at interfaces can be obtained; or based on risk records of similar roadways in coal mines, especially data obtained from anchor bolt measurements within roadways, statistical analysis of these data can yield the probability distribution of events. For example, there is a correlation between structural spalling and pressure; the probability of spalling can be obtained by analyzing the trend of pressure distribution. Furthermore, the distribution of events such as spalling and the displacement changes of the roof and floor slabs are similar, both being parameters affecting the attitude of the tunnel boring machine. The present invention obtains the information entropy of risk by tracking this data.
[0088] According to one embodiment of the present invention, the information entropy can be calculated in sections based on the working functional range of the tunnel boring machine, or it can be calculated as a whole. When calculating based on sections, if the information entropy of a region changes significantly and indicates a corresponding risk, it can be addressed by strengthening the support. For example, in the front section, in addition to the front shield, it can also be addressed by adding support and netting.
[0089] According to one embodiment of the present invention, by monitoring the deformation direction and deformation magnitude of the formed tunnel, the probability distribution of tunnel deformation is determined. When the pressure is approximate, the probability distribution assumes that the corresponding events have a consistent probability and is used to track the changes in the attitude of the tunnel boring machine.
[0090] Based on this, the beneficial effect of the present invention is that it solves the problem that the attitude change of the tunnel boring machine cannot be known through traditional technology.
[0091] 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.
[0092] 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.
[0093] 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.
[0094] 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.
[0095] 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.
[0096] 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.
[0097] 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.
[0098] 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 of attitude monitoring in a TBM excavation process, characterized by, include: When the tunnel excavation length is shorter than the threshold, the real-time parameters of the tunnel and the working parameters of the tunnel boring machine are obtained. When the tunnel excavation length first exceeds the threshold, a supervised offset prediction model is constructed. The input of the offset prediction model is the working parameters of the tunnel boring machine and the vibration of the cutterhead, and the output is the offset value of the cutterhead in the vertical direction. Based on the working parameters and acoustic signals of the tunnel boring machine during the tunneling process, it is determined whether a change in the tunnel category has occurred. When a change in tunnel category has occurred, the data of the new category is labeled and retrained. The trained offset prediction model is used to predict the offset of the cutterhead in the vertical direction. The information entropy during the tunneling process is calculated based on the predicted value, and the attitude is predicted based on the information entropy.
2. The attitude monitoring method in a TBM excavation process according to claim 1, wherein, The real-time parameters of the roadway include the pressure and displacement values of the roadway sidewalls, roof, and floor. The operating parameters of a tunnel boring machine (TBM) include the cutterhead's rotational speed, torque, power, cutterhead propulsion power, and thrust during the tunneling process; the values of the horizontal gyroscopes installed on the shield and each functional trolley section; and the pressure values at the TBM-tunnel contact surface.
3. A method of attitude monitoring in a TBM excavation process according to claim 2, wherein, The working parameters of the tunnel boring machine are used as input to the supervised offset prediction model to obtain the working condition category corresponding to the tunnel boring machine tunneling process; Clustering is performed based on the intensity of the acquired acoustic signals. The cluster labels are the operating condition categories. The cluster in which the operating condition belongs is determined based on the currently acquired acoustic signals, and whether it is consistent with the operating condition category is determined based on the distance from the cluster center.
4. The method of attitude monitoring in a TBM excavation process of claim 3, wherein, When the roadway category changes, manual operation is used instead of automatic operation. After a tunneling process is completed, the roadway category and working condition are marked. After the marking is completed, incremental training is carried out.
5. A method of attitude monitoring in a TBM excavation process as claimed in claim 4, wherein, When the tunnel excavation length first exceeds the threshold, and the cumulative excavation length of the autonomous driving system exceeds the autonomous driving threshold, an unsupervised offset prediction model is constructed. The input of the offset prediction model is the working parameters of the tunnel boring machine and the historical data of the cutterhead vibration. The output of the offset prediction model is the offset value of the tunnel boring machine cutterhead in space. Based on the output of the unsupervised offset prediction model, it is determined whether there is a need to switch to manual driving.
6. A method of attitude monitoring in a TBM excavation process as claimed in claim 5, wherein, Based on the mechanical model of similar roadways in the mining area, the response values of different parts of the roadway to pressure and facilities are obtained. The displacement probability of multiple parts in the shield tunnel is determined based on the mechanical response values. The information entropy is calculated based on the actual offset of each part in various events.
7. A method of attitude monitoring in a TBM excavation process as claimed in claim 6, wherein, When determining attitude changes based on information entropy, the support area is determined according to the location corresponding to the attitude change, and the support strategy for the corresponding part is changed.
8. A method of attitude monitoring in a TBM excavation process as claimed in claim 7, wherein, After the formation of different types of tunnels, monitoring data is obtained. The deformation direction and magnitude of the tunnels in response to sound vibration and tunneling time are statistically analyzed. The tunnel deformation probability distribution associated with shield parameters is obtained. Information entropy is calculated based on the probability distribution of tunnels of the same type. The difference in information entropy is used to determine whether the categories are consistent.
9. A method of attitude monitoring in a TBM excavation process according to claim 8, wherein, The vibration of each zone corresponding to the actual working parameters of the cutterhead is acquired, and the vibration frequency, vibration acceleration and vibration velocity of the effective wave of the shield cutterhead vibration signal are acquired. The vibration signal of the cutterhead is collected according to the different angles of the cutterhead corresponding to the corresponding zones and the real-time working conditions, and normalized using the Softmax function.
10. A system for attitude monitoring during TBM excavation, characterized by, include: Real-time parameter acquisition module: When the tunnel excavation length is shorter than the threshold, it acquires the instantaneous parameters of the tunnel and the working parameters of the tunnel boring machine. Offset prediction model building module: When the tunnel excavation length first exceeds the threshold, a supervised offset prediction model is built. The input of the offset prediction model is the working parameters of the tunnel boring machine and the vibration of the cutterhead, and the output is the offset value of the cutterhead in the vertical direction. Tunnel Category Confirmation Module: Based on the working parameters and acoustic signals of the tunnel boring machine during the tunneling process, determine whether a change in tunnel category has occurred. When a tunnel category shift occurs, label the data of the new category and retrain. Attitude change confirmation module: The trained offset prediction model is used to predict the offset of the cutterhead in the vertical direction, and the information entropy during the tunneling process is calculated based on the predicted value. Based on the information entropy, it is predicted whether the attitude has changed.
Citation Information
Patent Citations
Shield tunneling parameter deviation calculation method and system based on random forest regression
CN109918822A
Shield attitude and tunneling parameter multi-task prediction method
CN116595861A
