Geological Judgment Method, Device and Terminal Based on TBM Cutterhead Vibration Signal

Abstract

The present invention provides a geological judgment method, device and terminal based on the vibration signal of a TBM cutterhead. The method includes: acquiring the vibration signal, torque signal, thrust signal of the cutterhead during the current tunneling period and the spatial information of the TBM cutterhead; performing Fourier transform on the vibration signal to obtain the frequency spectrum of the vibration signal, and extracting the FBank features of the frequency spectrum; detecting abnormal points in the FBank features to determine the geological uniformity ratio of the current tunneling area corresponding to the current tunneling period; configuring a trained fusion model according to the geological uniformity ratio of the current tunneling area, and inputting the vibration signal, torque signal and thrust signal into the configured fusion model to obtain the rock mass grade of the current tunneling area; performing time series prediction based on the rock mass grade of the current tunneling area and the spatial information of the TBM cutterhead to obtain the geological judgment result. The present invention judges the current geological situation through the fusion model and then predicts the geological situation in the next period, so as to judge the geological situation during tunneling in advance.

Description

Technical Field

[0001] The present invention relates to the technical field of geological exploration, and particularly to a geological judgment method, device and terminal based on the vibration signal of a TBM cutter head. Background Technique

[0002] China has become the country with the largest scale and difficulty of tunnel (cave) construction in the world. A number of deep tunnels with remarkable characteristics such as "large buried depth, long tunnel line, complex geology, steep terrain, and frequent disasters" are being built or will be built in the fields of water conservancy and hydropower and transportation engineering. The Tunnel Boring Machine (TBM) has remarkable advantages such as "fast tunneling speed, high quality of the formed tunnel, high comprehensive economic benefits, and safe and civilized construction". However, the adaptability of TBM construction to bad geology is poor. Once encountering bad geology such as fault fracture zones, soft strata, karst, etc., disasters such as cave-ins, water inrush and mud inrush often occur. In the geological surveys carried out in the early stage of tunnel construction, the distance between survey points is usually far, and the geological conditions of each tunneling cycle cannot be judged in time, and operations are often carried out based on manual experience. Therefore, the phenomenon of abnormal damage of TBM is extremely likely to occur, which will cause serious problems such as construction period delay, economic losses and even casualties. Summary of the Invention

[0003] Embodiments of the present invention provide a geological judgment method, device and terminal based on the vibration signal of a TBM cutter head to solve the problem that the geological conditions during tunneling cannot be judged in time.

[0004] In a first aspect, embodiments of the present invention provide a geological judgment method based on the vibration signal of a TBM cutter head, including:

[0005] Obtain the vibration signal, torque signal, thrust signal of the cutter head and the spatial information of the TBM cutter head during the current tunneling period;

[0006] Perform Fourier transform on the vibration signal to obtain the spectrum of the vibration signal, and extract the FBank features of the spectrum;

[0007] Detect abnormal points of the FBank features, and determine the geological uniformity ratio of the current tunneling area corresponding to the current tunneling period based on the detection results;

[0008] Configure a trained fusion model according to the geological uniformity ratio of the current tunneling area, and input the vibration signal, torque signal and thrust signal into the configured fusion model to obtain the rock mass grade of the current tunneling area;

[0009] Based on the rock mass grade of the current tunneling area and the TBM cutterhead space information, perform time series prediction to obtain the rock mass grade of the next tunneling area and the TBM cutterhead space information corresponding to the next tunneling period, and use the rock mass grade and the TBM cutterhead space information of the next tunneling area as the geological judgment result; wherein, the next tunneling period is less than or equal to the current tunneling period.

[0010] In a possible implementation manner, input the vibration signal, torque signal, and thrust signal into the configured fusion model to obtain the rock mass grade of the current tunneling area, including:

[0011] Input the vibration signal, torque signal, and thrust signal corresponding to a preset angle range in the current tunneling area into the configured fusion model to obtain the rock mass grade corresponding to the preset angle range in the current tunneling area;

[0012] Based on the rock mass grades corresponding to multiple preset angle ranges in the current tunneling area, obtain the rock mass grade of the current tunneling area.

[0013] In a possible implementation manner, the fusion model includes a base learner for the vibration signal, a base learner for the thrust signal, a base learner for the torque signal, and a meta-learner respectively connected to the three base learners;

[0014] Input the vibration signal, torque signal, and thrust signal into the fusion model corresponding to the judgment result to obtain the rock mass grade of the current area, including:

[0015] Input the vibration signal, thrust signal, and torque signal into the corresponding base learners respectively, and perform cross-validation on each base learner to obtain multiple cross-validation results;

[0016] Vertically stack the respective verification results to obtain the prediction results of each base learner;

[0017] Horizontally splice the respective prediction results to obtain a feature matrix, and input the feature matrix into the meta-learner to obtain the rock mass grade of the current tunneling area output by the meta-learner.

[0018] In a possible implementation manner, before inputting the vibration signal, torque signal, and thrust signal into the fusion model corresponding to the judgment result to obtain the rock mass grade of the current tunneling area, the method further includes:

[0019] Establish an initial fusion model;

[0020] Obtain a training data set; the training data set includes multiple training samples, each training sample includes a vibration signal, a thrust signal, and a torque signal, and the label of each training sample is the rock mass grade;

[0021] Input each training sample into each base learner in the initial fusion model, and perform cross-validation on each base learner to obtain multiple cross-validation results;

[0022] Stack the validation results vertically to obtain the prediction results of each base learner;

[0023] Concatenate the prediction results horizontally to obtain a feature matrix;

[0024] Input the feature matrix into the meta-learner in the initial fusion model to train the meta-learner.

[0025] In a possible implementation, perform outlier detection on the FBank features, and determine the geological uniformity ratio of the current tunneling area corresponding to the current tunneling period based on the detection results, including:

[0026] If the fluctuation of the FBank features within a preset time period is less than or equal to a preset threshold, determine that the current area is a formation with uniform hardness and softness, and determine that the geological uniformity ratio of the current tunneling area corresponding to the current tunneling period is zero;

[0027] If the fluctuation of the FBank features within a preset time period is greater than the preset threshold, determine that the current area is a formation with uneven hardness and softness, and regard the FBank features with fluctuations greater than the preset threshold as abnormal FBank features;

[0028] Calculate the proportion of abnormal FBank features within a preset time period;

[0029] Configure the trained fusion model according to the geological uniformity ratio of the current tunneling area, including:

[0030] Adjust the weights of the trained fusion model based on the geological uniformity ratio of the current tunneling area.

[0031] In a possible implementation, extract the FBank features of the spectrum, including:

[0032] Obtain the filter frequency band interval, and set the filter based on the filter frequency band interval; the filter frequency band interval includes the filter frequency band interval of the concentrated frequency band and the filter frequency band interval of the non-concentrated frequency band;

[0033] Extract the FBank features of the spectrum through each filter;

[0034] Correspondingly, before extracting the FBank features of the spectrum, the method further includes:

[0035] Analyze the historical vibration frequency of the cutter head, and determine the concentrated frequency band and the non-concentrated frequency band of the historical vibration frequency;

[0036] Determine the filter intervals for the concentrated frequency bands and the non-concentrated frequency bands based on grid search and random forest algorithms.

[0037] In a second aspect, an embodiment of the present invention provides a geological judgment device based on the vibration signal of a TBM cutterhead, including:

[0038] An acquisition module, configured to acquire the vibration signal, torque signal, thrust signal of the cutterhead during the current tunneling period, and the spatial information of the TBM cutterhead;

[0039] A feature extraction module, configured to perform Fourier transform on the vibration signal to obtain the frequency spectrum of the vibration signal, and extract the FBank features of the frequency spectrum;

[0040] An anomaly detection module, configured to detect anomaly points in the FBank features, and determine the geological uniformity ratio of the current tunneling area corresponding to the current tunneling period based on the detection results;

[0041] A rock mass judgment module, configured to configure a trained fusion model according to the geological uniformity ratio of the current tunneling area, and input the vibration signal, torque signal, and thrust signal into the configured fusion model to obtain the rock mass grade of the current tunneling area;

[0042] A rock mass prediction module, configured to perform time series prediction based on the rock mass grade of the current tunneling area and the spatial information of the TBM cutterhead to obtain the rock mass grade and the spatial information of the TBM cutterhead of the next tunneling area corresponding to the next tunneling period, and use the rock mass grade and the spatial information of the next tunneling area as the geological judgment result; wherein, the next tunneling period is less than or equal to the current tunneling period.

[0043] In a third aspect, an embodiment of the present invention provides a terminal, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the method in the first aspect or any possible implementation manner of the first aspect are implemented.

[0044] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, the steps of the method in the first aspect or any possible implementation manner of the first aspect are implemented.

[0045] An embodiment of the present invention provides a geological judgment method based on the vibration signal of a TBM cutterhead, including: obtaining the vibration signal, torque signal, thrust signal of the cutterhead during the current tunneling period, and the spatial information of the TBM cutterhead; performing Fourier transform on the vibration signal to obtain the frequency spectrum of the vibration signal, and extracting the FBank features of the frequency spectrum; detecting abnormal points for the FBank features, and determining the geological uniformity ratio of the current tunneling area corresponding to the current tunneling period based on the detection results; configuring a trained fusion model according to the geological uniformity ratio of the current tunneling area, and inputting the vibration signal, torque signal and thrust signal into the configured fusion model to obtain the rock mass grade of the current tunneling area; performing time series prediction based on the rock mass grade of the current tunneling area and the spatial information of the TBM cutterhead to obtain the rock mass grade and the spatial information of the TBM cutterhead of the next tunneling area corresponding to the next tunneling period, and taking the rock mass grade and the spatial information of the TBM cutterhead of the next tunneling area as the geological judgment result; wherein, the next tunneling period is less than or equal to the current tunneling period. The present invention judges the current geological situation through the vibration signal, torque signal, thrust signal of the cutterhead during the current tunneling period and the fusion model, and then predicts the geological situation of the next period, so as to judge the geological situation during tunneling in advance. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for use in the embodiments or the description of the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0047] Figure 1 is the implementation flowchart of the geological judgment method based on the vibration signal of the TBM cutterhead provided by the embodiment of the present invention;

[0048] Figure 2 is the structural schematic diagram of the geological judgment device based on the vibration signal of the TBM cutterhead provided by the embodiment of the present invention;

[0049] Figure 3 is the schematic diagram of the terminal provided by the embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0050] In the following description, specific details such as specific system structures and technologies are presented for the purpose of illustration rather than limitation, so as to thoroughly understand the embodiments of the present invention. However, those skilled in the art should clearly understand that the present invention can also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid unnecessary details from interfering with the description of the present invention.

[0051] In order to make the purpose, technical solutions and advantages of the present invention more clear, specific embodiments will be described below in conjunction with the accompanying drawings.

[0052] See also Figure 1 , which shows a flow chart of the implementation of the geological judgment method based on the TBM cutterhead vibration signal provided by an embodiment of the present invention, which is described in detail as follows:

[0053] Step 101, obtaining the vibration signal, torque signal, thrust signal and TBM cutter head space information of the current excavation period.

[0054] In this embodiment, the current excavation period may include one or more excavation cycles, and the vibration signal, torque signal, and thrust signal of the cutterhead in the current excavation period may be acquired in real time through the sliding sampling window. The spatial information of the TBM cutterhead may be determined by the excavation mileage of the TBM excavator, and the spatial information is used to indicate the position of the rock mass.

[0055] In order to obtain the vibration signal, torque signal and thrust signal of the cutter disc, a vibration sensor, a gyroscope and a corresponding data transmission module can be set on the cutter disc in advance, wherein the vibration sensor can collect the vibration signal of the cutter disc, the gyroscope can collect the torque signal of the cutter disc, and the thrust signal of the cutter disc is obtained by the thrust detection sensor of the TBM. The data transmission module can transmit the vibration signal and torque signal of the cutter disc to the data analysis module, so that the data analysis module performs subsequent processing and analysis based on the vibration signal, torque signal and thrust signal of the cutter disc. Among them, a number of vibration sensors can be set on the periphery of the cutter disc of the TBM tunnel boring machine, and each vibration sensor can be evenly set on the surface of the cutter disc, or the vibration sensor can be set at half of the radius of the cutter disc and directly above the roller cutter to accurately sense the excitation signal of the roller cutter. The data analysis module can be set in the operation area of ​​the TBM tunnel boring machine.

[0056] Step 102, perform Fourier transform on the vibration signal to obtain the spectrum of the vibration signal, and extract the FBank feature of the spectrum.

[0057] In this embodiment, the spectrum can be smoothed by a Mel filter, and harmonic effects can be eliminated to highlight the common peak value and extract the FBank feature.

[0058] Step 103, perform anomaly detection on the FBank feature, and determine the geological uniform proportion of the current excavation area corresponding to the current excavation period based on the detection result.

[0059] In this embodiment, the object of abnormal point detection and analysis is the Fbank eigenvalue of the vibration signal, and the time length is one revolution of the cutter head. If the Fbank eigenvalue at a certain place within the circle is significantly different from that at other places, it indicates that the geology of the current tunneling area is a stratum with uneven hardness and softness. Moreover, according to the proportion of abnormal points, the unevenness degree of hardness and softness in the current tunneling area can be calculated, that is, the proportion of geological uniformity.

[0060] Step 104: Configure the trained fusion model according to the proportion of geological uniformity in the current tunneling area, and input the vibration signal, torque signal, and thrust signal into the configured fusion model to obtain the rock mass grade of the current tunneling area.

[0061] In this embodiment, the proportion of geological uniformity in the current tunneling area can, to a certain extent, reflect the high and low distribution of the rock mass grade in the current tunneling area. By setting the weights for analyzing the torque signal, thrust signal, and vibration signal in the fusion model based on the proportion of geological uniformity, the accuracy of the analysis results can be improved.

[0062] Step 105: Perform time series prediction based on the rock mass grade of the current tunneling area and the TBM cutter head spatial information to obtain the rock mass grade and TBM cutter head spatial information of the next tunneling area corresponding to the next tunneling period, and use the rock mass grade and TBM cutter head spatial information of the next tunneling area as the geological judgment result; wherein, the next tunneling period is less than or equal to the current tunneling period.

[0063] In this embodiment, the time series prediction is implemented based on a time series prediction model, and the time series prediction model can be a Long Short-Term Memory (LSTM) model. The specific process of time series prediction is as follows: In the tunneling direction, the three signals within each tunneling cycle section are input into the time series prediction model in chronological order, so as to obtain the rock mass grade of the next tunneling area output by the time series prediction model. For example: Extract the vibration signal, torque signal, and thrust signal features in fifty tunneling cycles according to the tunneling direction spatial information as input data, predict the rock mass grade of the area in the next tunneling cycle, and finally determine the spatial position of the soft surrounding rock in each angular range area in the next tunneling cycle, thereby realizing the prediction of the rock mass condition in front of the tunnel face. Among them, the fifty tunneling cycles in this embodiment are the current tunneling period, the next tunneling cycle is the next tunneling period, and the next tunneling period being less than or equal to the current tunneling period can make the prediction results more accurate.

[0064] The embodiment of the present invention determines the current geological situation through the vibration signal, torque signal, thrust signal of the cutter head in the current tunneling period and the fusion model, and then predicts the geological situation in the next period, which can judge the geological situation during tunneling in advance.

[0065] The present invention effectively identifies the vibration of the cutter head itself through the vibration sensor arranged at the cutter head of the TBM tunneling machine. The vibration sensor occupies a small space, has a simple structure, and is easy to implement. Through the data processing module, not only can the historical geological information of the TBM tunneling machine that has been tunnelled be inversely calculated, but also the geological information of the area to be tunnelled can be predicted, improving the construction efficiency of the project.

[0066] In a possible implementation manner, inputting the vibration signal, torque signal, and thrust signal into the configured fusion model to obtain the rock mass grade of the current tunneling area includes:

[0067] Inputting the vibration signal, torque signal, and thrust signal corresponding to the preset angle range in the current tunneling area into the configured fusion model to obtain the rock mass grade corresponding to the preset angle range in the current tunneling area;

[0068] Obtaining the rock mass grade of the current tunneling area based on the rock mass grades corresponding to multiple preset angle ranges in the current tunneling area.

[0069] In this embodiment, the gyroscope can also collect the real-time rotation angle of the cutter head. The real-time rotation angle of the cutter head belongs to the spatial information of the cutter head, and the vibration signal, torque signal, and thrust signal of the cutter head belong to the time signals of the cutter head. When the vibration signal, torque signal, and thrust signal of the cutter head are collected, these time domain signals should be added with spatial information, and the time domain signals are screened through the real-time rotation angle of the cutter head in the spatial information. When the rotation angle of the cutter head is within the preset angle range, the corresponding vibration signal, torque signal, and thrust signal are obtained. Inputting the screened vibration signal, torque signal, and thrust signal into the fusion model, the rock mass grade of the area corresponding to the preset angle range in the current tunneling area can be obtained.

[0070] For example, based on the spatial information provided by the gyroscope and the TBM tunneling mileage, using a sliding sampling window to add the 360° rotated by the cutter head within one tunneling cycle with spatial information, including: the azimuth angle (0° - 360°) of the cutter head and the tunneling mileage. The azimuth angle is detailedly divided into 36 parts, and then added to the vibration signal, torque signal, and thrust signal of the cutter head respectively, so as to obtain the vibration signal, torque signal, and thrust signal when the cutter head rotates through 0° - 10°, 10° - 20°... 350° - 360°. Then, the preset angle ranges are set as 0° - 10°, 10° - 20°... 350° - 360° respectively, and the vibration signal, torque signal, and thrust signal when the cutter head rotates through 0° - 10°, 10° - 20°... 350° - 360° are correspondingly input into the fusion model, and the rock mass grade when the cutter head rotates through 0° - 10°, 10° - 20°... 350° - 360° can be obtained. Based on the rock mass grades when the cutter head rotates through 0° - 10°, 10° - 20°... 350° - 360°, the rock mass grades of each subarea in the current tunneling area can be determined.

[0071] Based on this embodiment, the specific process of time series prediction is as follows: In the tunneling direction, three signals within a preset angle range in each tunneling cycle section are input into the time series prediction model in chronological order. For example, if the preset angle range is 0° - 10°, then according to the spatial information of the tunneling direction, the vibration signal, torque signal, and thrust signal features within the range of 0° - 10° of the cutterhead rotation angle in fifty tunneling cycles are extracted as input data. Based on the time series prediction method and the Long Short-Term Memory (LSTM) model, the rock mass grade in the area where the cutterhead rotation angle is within the range of 0° - 10° in the next tunneling cycle is predicted, and finally, the spatial positions of the weak surrounding rock in each angle range area in the next tunneling cycle are determined, thereby realizing the prediction of the rock mass condition in front of the tunnel face.

[0072] In a possible implementation, the fusion model includes a base learner for vibration signals, a base learner for thrust signals, a base learner for torque signals, and a meta-learner respectively connected to the three base learners;

[0073] Inputting the vibration signal, torque signal, and thrust signal into the fusion model corresponding to the judgment result to obtain the rock mass grade of the current area includes:

[0074] Inputting the vibration signal, thrust signal, and torque signal into the corresponding base learners respectively, and performing cross-validation on each base learner to obtain multiple cross-validation results;

[0075] Vertically stacking each verification result to obtain the prediction results of each base learner;

[0076] Horizontally splicing each prediction result to obtain a feature matrix, and inputting the feature matrix into the meta-learner to obtain the rock mass grade of the current tunneling area output by the meta-learner.

[0077] In this embodiment, the fusion model internally includes base learners for three types of signals. There are three base learners for each type of signal, namely Convolutional Neural Network (CNN), LSTM, and Light Gradient Boosting Machines (LGBM). Among them, for the three base learners of vibration signals, the input features are 20 groups of FBank frequency domain features of the current small-range vibration time series, as well as the maximum value, minimum value, peak-to-peak value, mean value, variance, root mean square, skewness, kurtosis, waveform factor, peak factor, impulse factor, and margin factor; for the three base learners of thrust signals, the input features are the maximum value, minimum value, peak-to-peak value, mean value, variance, root mean square, skewness, kurtosis, waveform factor, peak factor, impulse factor, and margin factor of the current small-range thrust time series; for the three base learners of torque signals, the input features are the maximum value, minimum value, peak-to-peak value, mean value, variance, root mean square, skewness, kurtosis, waveform factor, peak factor, impulse factor, and margin factor of the current small-range torque time series. After each base learner outputs the corresponding prediction result, the output data of each base learner are horizontally concatenated as the input data of the meta-learner, and the output data of the meta-learner is the rock mass grade. In a possible implementation manner, before inputting the vibration signal, torque signal, and thrust signal into the fusion model corresponding to the judgment result to obtain the rock mass grade of the current tunneling area, the method further includes:

[0078] Establish an initial fusion model;

[0079] Obtain a training data set; the training data set includes multiple training samples, each training sample includes a vibration signal, a thrust signal, and a torque signal, and the label of each training sample is the rock mass grade;

[0080] Input each training sample into each base learner in the initial fusion model, and perform cross-validation on each base learner to obtain multiple cross-validation results;

[0081] Vertically stack the validation results of each base learner to obtain the prediction results of each base learner;

[0082] Horizontally concatenate the prediction results of each base learner to obtain a feature matrix;

[0083] Input the feature matrix into the meta-learner in the initial fusion model to train the meta-learner.

[0084] In this embodiment, the specific steps for training the fusion model are as follows: First, the processed vibration, thrust, and torque signals are segmented into a training set and a test set. The samples in the training set are denoted as Mtrain, and the number of samples in the test set is denoted as Mtest. Then, the training set is input into each base learner, and cross-validation is performed on each base learner. On each base learner, the validation results of all cross-validations are stacked vertically to form a prediction result. The prediction results of all base learners are concatenated horizontally to form a feature matrix. The feature matrix is put into the meta-learner for training, and finally, three types of base learners corresponding to the three types of signals and the weights of the meta-learner corresponding to various uniformly distributed strata and strata with uneven hardness are obtained. Next, the test set is input into each base learner, and the corresponding results are predicted on each base learner. The prediction results of all base learners are concatenated horizontally to form a feature matrix, and the new feature matrix is put into the meta-learner for prediction.

[0085] In this embodiment, the samples in the training set may also include the vibration, thrust, and torque signals of the cutter head in the tunneling area with different geological uniform ratios. After training the fusion model using this training set, multiple corresponding relationships between the geological uniform ratio and the weights of the meta-learner in the fusion model can be obtained. Correspondingly, when using the fusion model, the weights of the meta-learner in the fusion model can be adjusted according to the detected geological uniform ratio.

[0086] In a possible implementation, for anomaly detection of FBank features, based on the detection results, determining the geological uniform ratio of the current tunneling area corresponding to the current tunneling period includes:

[0087] If the fluctuation of the FBank features within the preset time period is less than or equal to the preset threshold, it is determined that the current area is a stratum with uniform hardness, and the geological uniform ratio of the current tunneling area corresponding to the current tunneling period is determined to be zero;

[0088] If the fluctuation of the FBank features within the preset time period is greater than the preset threshold, it is determined that the current area is a stratum with uneven hardness, and the FBank features with fluctuations greater than the preset threshold are regarded as abnormal FBank features;

[0089] Calculate the proportion of abnormal FBank features within the preset time period;

[0090] Configuring the trained fusion model according to the geological uniform ratio of the current tunneling area includes:

[0091] Adjust the weights of the trained fusion model based on the geological uniform ratio of the current tunneling area.

[0092] In this embodiment, if a certain FBank eigenvalue is significantly different from other FBank eigenvalues, it indicates that there is a significant difference between the rock mass grade corresponding to this FBank eigenvalue and other rock mass grades, that is, this FBank eigenvalue reflects the existence of an area with uneven hardness in the current tunneling area.

[0093] In a possible implementation, extracting the FBank features of the spectrum includes:

[0094] Obtain the filter frequency band interval and set the filter based on the filter frequency band interval; the filter frequency band interval includes the filter frequency band interval of the concentrated frequency band and the filter frequency band interval of the non-concentrated frequency band;

[0095] Extract the FBank features of the spectrum through each filter;

[0096] Correspondingly, before extracting the FBank features of the spectrum, the method further includes:

[0097] Analyze the historical vibration frequency of the cutter head to determine the concentrated frequency band and non-concentrated frequency band of the historical vibration frequency;

[0098] Determine the filter interval of the concentrated frequency band and the filter interval of the non-concentrated frequency band based on the grid search and random forest algorithm.

[0099] In this embodiment, through analysis, the main frequency of the cutter head vibration is concentrated in 1 - 30Hz and 50 - 70Hz. It can be seen that these two frequency bands are of relatively high importance for prediction, and the remaining frequency bands are non-concentrated frequency bands. Therefore, the filter bank in the concentrated frequency band is encrypted, and the encryption interval is selected from 1, 2, 3, 4, 5Hz. The filters in the remaining frequency bands are in the non-encrypted area, and the filter interval is selected from 5, 10, 15Hz. Finally, test each filter interval on the random forest algorithm through grid search, and finally determine that the encryption interval is 3Hz and the non-encryption interval is 15Hz.

[0100] It should be understood that the magnitudes of the sequence numbers of the steps in the above embodiments do not mean 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 to the implementation process of the embodiments of the present invention.

[0101] The following is the device embodiment of the present invention. For the details not described in detail, reference can be made to the corresponding method embodiment above.

[0102] Figure 2 The structural schematic diagram of the geological judgment device based on the TBM cutter head vibration signal provided by the embodiment of the present invention is shown. For the sake of convenience of description, only the parts related to the embodiment of the present invention are shown and are described in detail as follows:

[0103] As Figure 2As shown, the geological judgment device 2 based on the TBM cutterhead vibration signal includes:

[0104] An acquisition module 21, configured to acquire the vibration signal, torque signal, thrust signal of the cutterhead during the current tunneling period, and the TBM cutterhead spatial information;

[0105] A feature extraction module 22, configured to perform Fourier transform on the vibration signal to obtain the frequency spectrum of the vibration signal, and extract the FBank features of the frequency spectrum;

[0106] An anomaly detection module 23, configured to detect anomaly points in the FBank features, and determine the geological uniformity ratio of the current tunneling area corresponding to the current tunneling period based on the detection results;

[0107] A rock mass judgment module 24, configured to configure a trained fusion model according to the geological uniformity ratio of the current tunneling area, and input the vibration signal, torque signal, and thrust signal into the configured fusion model to obtain the rock mass grade of the current tunneling area;

[0108] A rock mass prediction module 25, configured to perform time series prediction based on the rock mass grade of the current tunneling area and the TBM cutterhead spatial information to obtain the rock mass grade and TBM cutterhead spatial information of the next tunneling area corresponding to the next tunneling period, and use the rock mass grade and TBM cutterhead spatial information of the next tunneling area as the geological judgment result; wherein, the next tunneling period is less than or equal to the current tunneling period.

[0109] In a possible implementation manner, the rock mass judgment module 24 is specifically configured to:

[0110] Input the vibration signal, torque signal, and thrust signal corresponding to the preset angle range in the current tunneling area into the configured fusion model to obtain the rock mass grade corresponding to the preset angle range in the current tunneling area;

[0111] Obtain the rock mass grade of the current tunneling area based on the rock mass grades corresponding to multiple preset angle ranges in the current tunneling area.

[0112] In a possible implementation manner, the fusion model includes a base learner for the vibration signal, a base learner for the thrust signal, a base learner for the torque signal, and a meta-learner respectively connected to the three base learners;

[0113] The rock mass judgment module 24 is specifically configured to:

[0114] Input the vibration signal, thrust signal, and torque signal into the corresponding base learners respectively, and perform cross-validation on each base learner to obtain multiple cross-validation results;

[0115] Vertically stack the validation results of each to obtain the prediction results of each base learner;

[0116] Horizontally splice each prediction result to obtain a feature matrix, and input the feature matrix into the meta-learner to obtain the rock mass grade of the current tunneling area output by the meta-learner.

[0117] In a possible implementation, the geological judgment device 2 based on the TBM cutterhead vibration signal further includes:

[0118] A model training module, configured to establish an initial fusion model before inputting the vibration signal, torque signal, and thrust signal into the fusion model corresponding to the judgment result to obtain the rock mass grade of the current tunneling area;

[0119] Obtain a training data set; the training data set includes a plurality of training samples, each training sample includes a vibration signal, a thrust signal, and a torque signal, and the label of each training sample is the rock mass grade;

[0120] Input each training sample into each base learner in the initial fusion model, and perform cross-validation on each base learner to obtain a plurality of cross-validation results;

[0121] Vertically stack each validation result to obtain the prediction results of each base learner;

[0122] Horizontally splice each prediction result to obtain a feature matrix;

[0123] Input the feature matrix into the meta-learner in the initial fusion model to train the meta-learner.

[0124] In a possible implementation, the anomaly detection module 23 is specifically configured to:

[0125] If the FBank feature fluctuation within a preset time period is less than or equal to a preset threshold, determine that the current area is a formation with uniform hardness and softness, and determine that the geological uniformity ratio of the current tunneling area corresponding to the current tunneling period is zero;

[0126] If the FBank feature fluctuation within a preset time period is greater than the preset threshold, determine that the current area is a formation with uneven hardness and softness, and use the FBank feature with a fluctuation greater than the preset threshold as an abnormal FBank feature;

[0127] Calculate the proportion of the abnormal FBank feature within the preset time period;

[0128] Configure the trained fusion model according to the geological uniformity ratio of the current tunneling area, including:

[0129] Adjust the weight of the trained fusion model based on the geological uniformity ratio of the current tunneling area.

[0130] In a possible implementation, the feature extraction module 22 is specifically configured to:

[0131] Obtain the filter frequency band interval and set the filter based on the filter frequency band interval; the filter frequency band interval includes the filter frequency band interval of the concentrated frequency band and the filter frequency band interval of the non-concentrated frequency band;

[0132] Extract the FBank features of the spectrum through each filter;

[0133] Correspondingly, before extracting the FBank features of the spectrum, the method further includes:

[0134] Analyze the historical vibration frequency of the cutter head to determine the concentrated frequency band and the non-concentrated frequency band of the historical vibration frequency;

[0135] Determine the filter interval of the concentrated frequency band and the filter interval of the non-concentrated frequency band based on the grid search and the random forest algorithm.

[0136] Figure 3 It is a schematic diagram of the terminal provided by the embodiment of the present invention. As Figure 3 shown, the terminal 3 of this embodiment includes: a processor 30, a memory 31, and a computer program 32 stored in the memory 31 and executable on the processor 30. When the processor 30 executes the computer program 32, the steps in the above-mentioned embodiments of the geological judgment method based on the TBM cutter head vibration signal are implemented, such as Figure 2 the steps 101 to 105 shown. Alternatively, when the processor 30 executes the computer program 32, the functions of each module in the above-mentioned device embodiments are implemented, such as Figure 2 the functions of the modules 21 to 25 shown.

[0137] Exemplarily, the computer program 32 can be divided into one or more modules, and the one or more modules are stored in the memory 31 and executed by the processor 30 to complete the present invention. The one or more modules can be a series of computer program instruction segments capable of performing specific functions, and these instruction segments are used to describe the execution process of the computer program 32 in the terminal 3. For example, the computer program 32 can be divided into Figure 2 the modules 21 to 25 shown.

[0138] The terminal 3 can be a computing device such as a desktop computer, a notebook, a palm computer, and a cloud server. The terminal 3 may include, but is not limited to, a processor 30 and a memory 31. Those skilled in the art can understand that Figure 3 merely examples of the terminal 3 do not constitute a limitation on the terminal 3, and it may include more or fewer components than shown in the figure, or combine certain components, or different components. For example, the terminal may further include input and output devices, network access devices, a bus, etc.

[0139] The so-called processor 30 may be a Central Processing Unit (CPU), or may also be other general-purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field-Programmable Gate Arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0140] The memory 31 may be an internal storage unit of the terminal 3, such as the hard disk or memory of the terminal 3. The memory 31 may also be an external storage device of the terminal 3, such as a plug-in hard disk equipped on the terminal 3, a Smart Media Card (SMC), a Secure Digital (SD) card, a Flash Card, etc. Further, the memory 31 may also include both the internal storage unit of the terminal 3 and the external storage device. The memory 31 is used to store the computer program and other programs and data required by the terminal. The memory 31 may also be used to temporarily store data that has been output or will be output.

[0141] Those skilled in the art can clearly understand that, for the convenience and conciseness of description, only the above division of each functional unit and module is used as an example. In practical applications, the above functions can be allocated to different functional units and modules according to needs, that is, the internal structure of the device is divided into different functional units or modules to complete all or part of the functions described above. Each functional unit and module in the embodiment may be integrated into one processing unit, or each unit may exist physically alone, or two or more units may be integrated into one unit. The above integrated unit may be implemented in the form of hardware or in the form of a software functional unit. In addition, the specific names of each functional unit and module are only for the convenience of mutual distinction and do not limit the protection scope of the present application. The specific working processes of the units and modules in the above system can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0142] In the above embodiments, the descriptions of the various embodiments have their own emphases. For the parts not detailed or recorded in a certain embodiment, reference may be made to the relevant descriptions of other embodiments.

[0143] Those of ordinary skill in the art will realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in hardware or software depends on the specific application and design constraints of the technical solution. A professional technician can use different methods for each specific application to implement the described functions, but such implementation should not be considered to exceed the scope of the present invention.

[0144] In the embodiments provided by the present invention, it should be understood that the disclosed device / terminal and method can be implemented in other ways. For example, the device / terminal embodiments described above are merely illustrative. For example, the division of the modules or units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed coupling or direct coupling or communication connection to each other can be through some interfaces. The indirect coupling or communication connection of the device or unit can be in electrical, mechanical or other forms.

[0145] The units described as separate components may or may not be physically separated. The components displayed as units may or may not be physical units, that is, they can be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0146] In addition, the functional units in each embodiment of the present invention can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit. The above integrated units can be implemented in the form of hardware or in the form of software functional units.

[0147] When the integrated module is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, to implement all or part of the processes in the above-mentioned embodiment methods of the present invention, it can also be completed by instructing relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by a processor, the steps of the above-mentioned various embodiments of the geological judgment method based on the TBM cutterhead vibration signal can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in the form of source code, object code, executable file or some intermediate form, etc. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disc, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal, and software distribution medium, etc. It should be noted that the content included in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, the computer-readable medium does not include electrical carrier signals and telecommunication signals.

[0148] The above-mentioned embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the various embodiments of the present invention, and should all be included in the protection scope of the present invention.

Claims

1. A geological judgment method based on the vibration signal of the TBM cutterhead, characterized in that Including: Obtain the vibration signal, torque signal, thrust signal of the cutter head during the current tunneling period, and the spatial information of the TBM cutter head; Perform Fourier transform on the vibration signal to obtain the frequency spectrum of the vibration signal, and extract the FBank features of the frequency spectrum; Detect abnormal points in the FBank features, and determine the geological uniformity ratio of the current tunneling area corresponding to the current tunneling period based on the detection results; The detecting abnormal points in the FBank features and determining the geological uniformity ratio of the current tunneling area corresponding to the current tunneling period based on the detection results includes: Calculate the proportion of abnormal points to obtain the geological uniformity ratio of the current tunneling area; Configure the trained fusion model according to the geological uniformity ratio of the current tunneling area, and input the vibration signal, the torque signal, and the thrust signal into the configured fusion model to obtain the rock mass grade of the current tunneling area; The fusion model includes a base learner for the vibration signal, a base learner for the thrust signal, a base learner for the torque signal, and a meta-learner respectively connected to the three base learners; The configuring the trained fusion model according to the geological uniformity ratio of the current tunneling area includes: Based on the corresponding relationship between multiple geological uniformity ratios and the weights of the meta-learner in the fusion model, determine the weights of the meta-learner corresponding to the geological uniformity ratio of the current tunneling area, and adjust the weights in the trained fusion model accordingly; Perform time series prediction based on the rock mass grade and the TBM cutter head spatial information of the current tunneling area to obtain the rock mass grade and the TBM cutter head spatial information of the next tunneling area corresponding to the next tunneling period, and use the rock mass grade and the TBM cutter head spatial information of the next tunneling area as the geological judgment result; wherein, the next tunneling period is less than or equal to the current tunneling period.

2. The geological judgment method based on the TBM cutter head vibration signal according to claim 1, characterized in that Inputting the vibration signal, the torque signal, and the thrust signal into the configured fusion model to obtain the rock mass grade of the current tunneling area includes: Input the vibration signal, torque signal, and thrust signal corresponding to the preset angle range in the current tunneling area into the configured fusion model to obtain the rock mass grade corresponding to the preset angle range in the current tunneling area; Obtain the rock mass grade of the current tunneling area based on the rock mass grades corresponding to multiple preset angle ranges in the current tunneling area.

3. The geological judgment method based on the TBM cutter head vibration signal according to claim 1, wherein Inputting the vibration signal, the torque signal, and the thrust signal into the configured fusion model to obtain the rock mass grade of the current tunneling area includes: Input the vibration signal, the thrust signal, and the torque signal into the corresponding base learners respectively, and perform cross-validation on each base learner to obtain multiple cross-validation results; Vertically stack the verification results of each to obtain the prediction results of each base learner; Horizontally splice the prediction results of each to obtain a feature matrix, and input the feature matrix into the meta-learner to obtain the rock mass grade of the current tunneling area output by the meta-learner.

4. The geological judgment method based on the TBM cutterhead vibration signal according to claim 3, wherein Before inputting the vibration signal, the torque signal, and the thrust signal into the configured fusion model to obtain the rock mass grade of the current tunneling area, the method further includes: Establish an initial fusion model; Obtain a training data set; the training data set includes a plurality of training samples, each training sample includes a vibration signal, a thrust signal, and a torque signal, and the label of each training sample is the rock mass grade; Input each training sample into each base learner in the initial fusion model, and perform cross-validation on each base learner to obtain multiple cross-validation results; Vertically stack the respective verification results to obtain the prediction results of each base learner; Horizontally splice the respective prediction results to obtain a feature matrix; Input the feature matrix into the meta-learner in the initial fusion model to train the meta-learner.

5. The geological judgment method based on the TBM cutterhead vibration signal according to claim 1, characterized in that The abnormal point detection of the FBank feature and determining the geological uniformity ratio of the current tunneling area corresponding to the current tunneling period based on the detection result includes: If the fluctuation of the FBank feature within a preset time period is less than or equal to a preset threshold, it is determined that the current area is a stratum with uniform hardness and softness, and the geological uniformity ratio of the current tunneling area corresponding to the current tunneling period is determined to be zero; If the fluctuation of the FBank feature within a preset time period is greater than the preset threshold, it is determined that the current area is a stratum with uneven hardness and softness, and the FBank feature with a fluctuation greater than the preset threshold is used as an abnormal FBank feature; Calculate the proportion of the abnormal FBank feature within the preset time period.

6. The geological judgment method based on the TBM cutterhead vibration signal according to claim 1, wherein The extraction of the FBank feature of the spectrum includes: Obtain the filter frequency band interval, and set filters based on the filter frequency band interval; the filter frequency band interval includes the filter frequency band interval of the concentrated frequency band and the filter frequency band interval of the non-concentrated frequency band; Extract the FBank feature of the spectrum through each filter; Correspondingly, before the extraction of the FBank feature of the spectrum, the method further includes: Analyze the historical vibration frequency of the cutterhead to determine the concentrated frequency band and the non-concentrated frequency band of the historical vibration frequency; Determine the filter interval of the concentrated frequency band and the filter interval of the non-concentrated frequency band based on the grid search and random forest algorithm.

7. A geological judgment device based on the vibration signal of a TBM cutterhead, characterized in that, Includes: An acquisition module, configured to acquire the vibration signal, the torque signal, the thrust signal of the cutterhead during the current tunneling period, and the spatial information of the TBM cutterhead; A feature extraction module, configured to perform Fourier transform on the vibration signal to obtain the spectrum of the vibration signal, and extract the FBank feature of the spectrum; An abnormal detection module, configured to perform abnormal point detection on the FBank feature, and determine the geological uniformity ratio of the current tunneling area corresponding to the current tunneling period based on the detection result; The abnormal detection module is specifically configured to: Calculate the proportion of abnormal points to obtain the geological uniformity ratio of the current tunneling area; A rock mass judgment module, configured to configure a trained fusion model according to the geological uniformity ratio of the current tunneling area, and input the vibration signal, the torque signal, and the thrust signal into the configured fusion model to obtain the rock mass grade of the current tunneling area; The fusion model includes a base learner for vibration signals, a base learner for thrust signals, a base learner for torque signals, and a meta-learner respectively connected to the three base learners; The rock mass judgment module is specifically used for: Based on the corresponding relationship between multiple geological uniformity ratios and the weights of the meta-learner in the fusion model, determine the geological uniformity ratio of the current tunneling area, determine the corresponding weights of the meta-learner, and accordingly adjust the weights in the trained fusion model; A rock mass prediction module, which is used to perform time series prediction based on the rock mass grade and TBM cutterhead spatial information of the current tunneling area, obtain the rock mass grade and TBM cutterhead spatial information of the next tunneling area corresponding to the next tunneling period, and use the rock mass grade and TBM cutterhead spatial information of the next tunneling area as the geological judgment result; wherein, the next tunneling period is less than or equal to the current tunneling period.

8. The geological judgment device based on the TBM cutterhead vibration signal according to claim 7, characterized in that, The rock mass judgment module is specifically used for: Input the vibration signals, torque signals, and thrust signals corresponding to the preset angle range in the current tunneling area into the configured fusion model to obtain the rock mass grade corresponding to the preset angle range in the current tunneling area; Obtain the rock mass grade of the current tunneling area based on the rock mass grades corresponding to multiple preset angle ranges in the current tunneling area.

9. A terminal, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the geological judgment method based on the TBM cutterhead vibration signal according to any one of claims 1 to 6 above.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the geological judgment method based on the TBM cutterhead vibration signal according to any one of claims 1 to 6 above.

Images