Rock mass condition prediction method and device, equipment and storage medium

Abstract

The present application provides a kind of rock mass condition prediction method and device, equipment and storage medium. Including: obtaining the original tunneling data of tunnel boring machine and the rock mass state grade corresponding to original tunneling data, as original data;According to original data, feature screening and feature construction are carried out, and model training data set is formed;According to model training data set, multiple machine learning model training algorithms are used for model training, in the process of model training, the northern hawk optimization algorithm improved by longitudinal and transverse cross strategy is used to optimize the hyperparameters in the process of model training;According to the model performance of the multiple machine learning models obtained by training, the machine learning model with the best model performance is used as the rock mass condition prediction model;The original tunneling data of tunnel boring machine is obtained;The original tunneling data is subjected to feature screening and feature construction to form a tunneling data set;The tunneling data set is input into the rock mass condition prediction model to predict the rock mass condition.

Description

Technical Field

[0001] This invention belongs to the field of TBM intelligent sensing construction technology, and particularly relates to a method, device, equipment and storage medium for predicting rock mass conditions. Background Technology

[0002] Compared to drill-and-blast methods, tunnel boring machine (TBM) construction offers advantages such as high tunneling efficiency, high productivity, environmental friendliness, and minimal impact on the surrounding environment. However, when tunneling in complex and structurally variable rock formations, TBMs are highly susceptible to accidents such as jamming or equipment damage due to large rock deformations. Therefore, real-time monitoring and understanding of tunnel rock conditions are crucial for construction safety, strengthening support systems, and reducing construction risks.

[0003] Existing rock mass condition information mostly comes from pre-construction borehole exploration, advance drilling during construction, and geophysical exploration methods such as TRT or TSP. These methods require specialized detection equipment and professional personnel, and in most cases, the TBM cannot advance normally during the exploration process, which can affect the progress of the project to some extent. Therefore, it is urgent to provide a more convenient method for predicting rock mass conditions. Summary of the Invention

[0004] This invention provides a method, apparatus, equipment, and storage medium for predicting rock mass conditions, which can form a rock mass condition prediction model with high prediction accuracy.

[0005] On the one hand, a method for predicting rock mass conditions is provided, including:

[0006] Obtain the raw tunneling data of the tunnel boring machine and the corresponding rock mass condition level as the raw data;

[0007] Based on the raw data, feature selection and feature construction are performed to form a model training dataset;

[0008] Based on the model training dataset, multiple machine learning model training algorithms were used for model training. During the model training process, the Northern Eagle optimization algorithm, which was improved through a cross-hatching strategy, was used to optimize the hyperparameters during the model training process.

[0009] Based on the model performance of multiple machine learning models obtained from training, the machine learning model with the best model performance is selected as the rock mass condition prediction model.

[0010] Obtain the raw tunneling data from the tunnel boring machine;

[0011] The raw tunneling data from the tunnel boring machine is subjected to feature filtering and feature construction to form a tunneling dataset;

[0012] The tunneling dataset is input into the rock mass condition prediction model to predict the rock mass conditions during the tunneling process.

[0013] Optionally, feature selection and feature construction are performed on the original data to form a model training dataset, including:

[0014] The Isolation Forest algorithm is used to remove outliers from the original data and fill in the null values ​​in the original data to form the first tunneling parameter dataset;

[0015] Feature filtering is performed on the first tunneling parameter dataset. The variance of a single feature and the correlation between features are calculated. Features with variances below the variance threshold or correlations above the correlation threshold are removed to obtain the second tunneling parameter dataset.

[0016] Based on the data in the first tunneling parameter dataset, feature construction is performed to obtain a tunneling performance feature dataset, which includes thrust penetration index, torque penetration index and specific penetration.

[0017] The second tunneling parameter dataset and the tunneling performance feature dataset are merged to obtain the model training dataset.

[0018] Optionally, based on the data in the first tunneling parameter dataset, feature construction is performed to obtain a tunneling performance feature dataset, which includes thrust penetration index, torque penetration index, and specific penetration, including:

[0019] According to the formula Feature construction was performed to obtain the field thrust penetration index. The Indicates thrust, the Indicates the number of hobs, the Indicates penetration degree;

[0020] According to the formula Feature construction was performed to obtain the field torque penetration index. The Indicates the torque of the cutter head, the Indicates the diameter of the cutter head;

[0021] According to the formula Feature construction is performed to obtain the penetration degree. .

[0022] Optionally, during model training, the improved Northern Eagle optimization algorithm, modified through a cross-cutting strategy, is used to optimize the hyperparameters during model training, including:

[0023] Step 1, Initialize the Northern Goshawk population , The This represents the population matrix of the Northern Goshawk. Indicates the first The location of the northern goshawk, the aforementioned Indicates the first The first Northern Eagle The position of each dimension of the Northern Goshawk represents the value of a hyperparameter.

[0024] Step 2: Determine the value of a certain model performance index of the machine learning model corresponding to the Northern Goshawk based on the location of the Northern Goshawk, and use it as the fitness value; the model performance index includes any one of the following: the weighted average of accuracy and recall, accuracy, and the area under the receiver operating feature curve.

[0025] Step 3, use the formula , , To identify and attack prey; Indicates the first After the first Northern Goshawk identified and attacked its prey, the second... Wei's new position; Indicates the first Fitness value of a Northern Goshawk; Indicates the first The location of the prey of the northern eagle; Indicates the first The first of the prey of the northern goshawk The position of the dimension; Indicates the first The fitness value of the prey of the Northern Goshawk; Indicates the first Target location update during prey identification and attack operations by a Northern Goshawk; Indicates the first step in prey identification and attack operations. The fitness value of the target location of a Northern Goshawk is updated. for Random numbers within; It is a random number, either 1 or 2;

[0026] Step 4, use the formula , , They conducted chase and escape maneuvers; among them, Indicates the first After the Northern Goshawk completed its pursuit and escape maneuvers, the... Wei's new position; Indicates the first The target location is updated when a Northern Goshawk performs a chase and escape maneuver. Indicates the first The fitness value of the target's updated position when a Northern Goshawk performs a chase and escape maneuver; Indicates the current iteration number; Indicates the maximum number of iterations;

[0027] Step 5: Perform a horizontal cross operation between the position of the Northern Goshawk obtained in Step 2 and the position of the Northern Goshawk obtained in Step 4.

[0028] Step 6: Perform a vertical cross-operation on the position of the Northern Goshawk in Step 2;

[0029] Step 7: Compare the fitness values ​​of the Northern Goshawks before and after the longitudinal crossing in Step 6, and retain the Northern Goshawks with higher fitness values.

[0030] Step 8: Repeat steps 2 through 7 until the maximum number of iterations is met. At that time, the coordinates of the Northern Goshawk with the highest fitness value are output as the hyperparameter combination of the machine learning model.

[0031] Optionally, the positions of the Northern Goshawks obtained in step 2 and step 4 can be horizontally intersected, including:

[0032] According to the formula , The location of the Northern Goshawk in step 2 The location of the Northern Goshawk obtained in step 4 The Dimensions intersect horizontally; among them, This indicates the first step in step 2. The first Northern Eagle Dimensional position; This indicates the result obtained in step 4. The first Northern Eagle Dimensional position; This indicates the first step in step 2. The offspring of the Northern Goshawk obtained by lateral crossbreeding dimension; This indicates the result obtained in step 4. The offspring of the Northern Goshawk obtained by lateral crossbreeding dimension; and for A random number between [a certain number of points].

[0033] Optionally, a vertical cross-operation is performed on the position of the Northern Goshawk in step 2, including:

[0034] According to the formula The Northern Goshawk in step 2 The peacekeeping Dimensions are vertically intersected; among them, To bring the Northern Goshawk in step 2 The peacekeeping The offspring individuals obtained after vertical crossover of dimensions dimension.

[0035] Optionally, based on the model performance of multiple trained machine learning models, the machine learning model with the best performance is selected as the rock mass condition prediction model, including:

[0036] The weighted average of the accuracy and recall of the trained machine learning model, the area under the curve of the accuracy and receiver operating feature curve are summed to obtain the sum of the model performance metrics of the machine learning model.

[0037] The machine learning model that accumulates the highest sum of the model performance indicators is used as the rock mass condition prediction model.

[0038] On the other hand, a rock mass condition prediction device is provided, comprising:

[0039] The acquisition module is used to acquire the raw tunneling data of the TBM and the corresponding rock mass condition level as the raw data;

[0040] The data processing module is used to perform feature filtering and feature construction based on the raw data to form a model training dataset;

[0041] The model training module is used to train models using multiple machine learning model training algorithms based on the model training dataset. During the model training process, the Northern Eagle optimization algorithm, which is improved through a cross-cutting strategy, is used to optimize the hyperparameters of the model training process. Based on the model performance of the multiple machine learning models obtained from the training, the machine learning model with the best performance is selected as the rock mass condition prediction model.

[0042] On the other hand, an electronic device is provided, including the rock mass condition prediction device as described above.

[0043] On the other hand, a computer-readable storage medium is provided, wherein at least one piece of program code is stored in the computer-readable storage medium, the program code being executed by a processor to implement the rock mass condition prediction method as described in any of the preceding claims.

[0044] The beneficial effects of the technical solutions provided in this disclosure are:

[0045] This disclosure provides a method for predicting rock mass conditions. It obtains raw tunneling data from a tunnel boring machine and the corresponding rock mass condition levels as raw data. Feature filtering and feature construction are performed on the raw data to form a model training dataset. The data in the model training dataset is more conducive to the performance of the trained machine learning model compared to the raw data. During model training, the Northern Eagle optimization algorithm, improved with a cross-cutting strategy, is used to optimize the hyperparameters, ensuring that the hyperparameter combination used in the machine learning model training is optimal, thereby further improving the model performance. Finally, by evaluating the model performance metrics, the model with the best performance is selected as the rock mass condition prediction model, ensuring the model's prediction accuracy. Attached Figure Description

[0046] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0047] Figure 1 A flowchart of a rock mass condition prediction method provided in this embodiment of the disclosure;

[0048] Figure 2 A complete cycle diagram of TBM tunneling operation parameters is provided in this embodiment of the disclosure;

[0049] Figure 3 A flowchart of a raw data processing procedure provided in this embodiment of the disclosure;

[0050] Figure 4 A flowchart of model training provided for embodiments of this disclosure;

[0051] Figure 5 A structural block diagram of a rock mass condition prediction device provided in this embodiment of the present disclosure;

[0052] Figure 6 This is a structural block diagram of an electronic device provided in an embodiment of the present disclosure.

[0053] The attached figures are labeled as follows:

[0054] 21: Acquisition Module; 22: Data Processing Module; 23: Model Building Module;

[0055] 31: Processor; 32: Memory. Detailed Implementation

[0056] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0057] Essentially, the real-time tunneling operation parameters generated by the TBM are the result of the interaction between the TBM and the rock mass. Therefore, these parameters can characterize the current tunneling status of the TBM and indirectly reflect the geological conditions of the surrounding rock. Further analysis of these TBM operation parameters and the establishment of machine learning models based on them can predict rock mass conditions in real time. This facilitates real-time decision-making by TBM operators and construction personnel, enabling them to select appropriate support and control measures, reduce construction risks, and is of great significance for the safe construction of TBM tunnel projects.

[0058] Figure 1 A flowchart illustrating a rock mass condition prediction method provided in this embodiment of the disclosure. See also... Figure 1 ,include:

[0059] S11. Obtain the original tunneling data of the TBM and the corresponding rock mass condition level as the original data.

[0060] Raw tunneling data for tunnel boring machines (TBMs) includes key tunneling operating parameters (cutterhead speed, penetration depth, thrust, torque, etc.), mechanical motor current and output power parameters, mechanical pressure and stress parameters, and mechanical oil level and temperature. Different TBM models will produce different raw tunneling data.

[0061] In this embodiment of the disclosure, regarding the rock mass condition level, for highway tunnels excavated by TBM, the rock mass condition level is divided into six levels: I, II, III, IV, V, and VI, based on the rock hardness and integrity and with reference to the "Design Code for Highway Tunnels" (JTG 3370.1-2018); for water conveyance tunnels excavated by TBM, the rock mass condition level is divided into five categories: I, II, III, IV, and V, based on rock strength, integrity, structural surface condition, groundwater and structural surface occurrence and with reference to the "Code for Geological Investigation of Water Conservancy and Hydropower Projects" (GB 50487-2008).

[0062] S12. Based on the original data, perform feature selection and feature construction to form a model training dataset.

[0063] In one example, step S12 includes:

[0064] Step 1: Use the Isolation Forest algorithm to remove outliers from the original data and fill in the null values ​​in the original data to form the first tunneling parameter dataset.

[0065] In this embodiment of the disclosure, the steps of the isolated forest method are as follows:

[0066] The first step is to select multiple groups of small-volume samples.

[0067] The second step is to construct an isolated tree for each group of samples, randomly select a feature at each node, and randomly select a split value between the maximum and minimum values ​​of the feature. Based on the split value, the samples are divided into the left and right subtrees. The termination condition is when the tree reaches a limited height or the number of samples at the node reaches a certain number.

[0068] The third step is to repeat the first and second steps to build a specific number of isolated trees, forming an isolated forest.

[0069] The fourth step involves calculating the average path length required for a sample to travel from the root node to the isolated node within the isolated forest. Samples with shorter average path lengths are identified as outliers and removed. Furthermore, to fill in missing values ​​in the recorded features, the average value of the feature variables from adjacent time periods is selected, ultimately forming the first tunneling parameter dataset.

[0070] Step 2: Perform feature filtering on the first tunneling parameter dataset, calculate the variance of a single feature and the correlation between features, and remove features with variances below the variance threshold or correlations above the correlation threshold to obtain the second tunneling parameter dataset.

[0071] In one example, step 2 includes:

[0072] The first step is to follow the formula. The variance of a single feature is calculated, where, This indicates the number of data points involved in the calculation for a particular feature. A single data point representing a certain characteristic. It represents the mean of the data used in the calculation for a certain feature.

[0073] The second step is to calculate the Pearson correlation coefficient using the formula. Calculate the correlation between features. and These are data variables representing two features (i.e., one of the data used in the calculation for the features). and It represents the average value of the data variable (that is, the average value of the data in which the feature is used in the calculation).

[0074] The third step is to remove features with low variance (variance below the variance threshold) and features with high correlation (correlation above the correlation threshold).

[0075] For example, the variance threshold can be 1.5 and the correlation threshold can be 0.8.

[0076] Of course, the above threshold is only one example provided by the embodiments of this disclosure, and variance threshold and correlation threshold can be selected selectively, which is not limited by this disclosure.

[0077] Step 3: Based on the data in the first tunneling parameter dataset, feature construction is performed to obtain a tunneling performance feature dataset, which includes thrust penetration index, torque penetration index, and specific penetration.

[0078] In one example, step 3 includes:

[0079] The first step is to follow the formula. Feature construction was performed to obtain the field thrust penetration index. The Indicates thrust, the Indicates the number of hobs, the Indicates the degree of penetration.

[0080] The second step is to follow the formula. Feature construction was performed to obtain the field torque penetration index. The Indicates the torque of the cutter head, the Indicates the diameter of the cutter head.

[0081] Third step, according to the formula Feature construction is performed to obtain the penetration degree. .

[0082] In this embodiment of the disclosure, features are constructed using the above formula, and calculated based on the original data. , and This can describe the tunneling performance characteristics of tunnel boring machines, improve the richness of training data, and help improve the prediction accuracy of the trained model.

[0083] Step 4: Merge the second tunneling parameter dataset and the tunneling performance feature dataset to obtain the model training dataset.

[0084] In this embodiment of the disclosure, by performing outlier removal, null value imputation, feature filtering, and feature construction on the original data, the integrity of the original data can be ensured on the one hand, and the training data can be ensured to use better data, which is beneficial to the prediction accuracy of the trained model.

[0085] S13. Based on the model training dataset, multiple machine learning model training algorithms are used to train the model. During the model training process, the improved Northern Eagle optimization algorithm with cross-hatching strategy is used to optimize the hyperparameters during the model training process.

[0086] In one example, step S13 includes:

[0087] Step 1, Initialize the Northern Goshawk population , The This represents the population matrix of the Northern Goshawk. Indicates the first The location of the northern goshawk, the aforementioned Indicates the first The first Northern Eagle The position of each dimension of the Northern Eagle represents the value of a hyperparameter.

[0088] Step 2: Determine the value of a certain model performance index of the machine learning model corresponding to the Northern Goshawk based on the location of the Northern Goshawk, and use it as the fitness value. The model performance index includes any one of the following: the weighted average of accuracy and recall (F1), accuracy (Acurracy), and the area under the receiver operating feature curve (AUC).

[0089] Step 3, use the formula , , To identify and attack prey; Indicates the first After the first Northern Goshawk identified and attacked its prey, the second... Wei's new position; Indicates the first Fitness value of a Northern Goshawk; Indicates the first The location of the prey of the northern eagle; Indicates the first The first of the prey of the northern goshawk The position of the dimension; Indicates the first The fitness value of the prey of the Northern Goshawk; Indicates the first Target location update during prey identification and attack operations by a Northern Goshawk; Indicates the first step in prey identification and attack operations. The fitness value of the target location of a Northern Goshawk is updated. for Random numbers within; It is a random number, either 1 or 2.

[0090] Step 4, use the formula , , They conducted chase and escape maneuvers; among them, Indicates the first After the Northern Goshawk completed its pursuit and escape maneuvers, the... Wei's new position; Indicates the first The target location is updated when a Northern Goshawk performs a chase and escape maneuver. Indicates the first The fitness value of the target's updated position when a Northern Goshawk performs a chase and escape maneuver; Indicates the current iteration number; This indicates the maximum number of iterations.

[0091] Step 5: Perform a horizontal cross operation between the position of the Northern Goshawk obtained in Step 2 and the position of the Northern Goshawk obtained in Step 4.

[0092] In one example, step 5 includes:

[0093] According to the formula , Parental individuals and the offspring individuals formed in step 4 The Dimensions intersect horizontally; among them, and Representing the parent individuals and offspring individuals The dimension, and Representing the parent individuals Offspring individuals The first generation of offspring obtained by lateral crossover dimension, and for A random number between [a certain number of points].

[0094] In this embodiment of the disclosure, the above formula can be used to achieve horizontal cross-operation between parent and offspring individuals, further enhancing the richness of the Northern Goshawk population.

[0095] Step 6: Perform a vertical cross operation on the position of the Northern Eagle in Step 2.

[0096] In one example, step 6 includes:

[0097] According to the formula Parental individuals The peacekeeping Dimensions are vertically intersected, where, For the parent generation The peacekeeping The offspring obtained after vertical crossover of dimensions dimension.

[0098] In this embodiment of the disclosure, the richness of the population is enhanced by performing a longitudinal crossover operation on the parent individuals of the Northern Goshawk using the above formula.

[0099] Step 7: Compare the fitness values ​​of the Northern Goshawks before and after the longitudinal crossing in Step 6, and retain the Northern Goshawks with higher fitness values.

[0100] Step 8: Repeat steps 2 through 7 until the maximum number of iterations is met. At that time, the coordinates of the Northern Goshawk with the highest fitness value are output as the hyperparameter combination of the machine learning model.

[0101] In this embodiment of the disclosure, by using the cross-sectional strategy to improve and optimize the Northern Eagle optimization algorithm, the ability of the Northern Eagle algorithm to escape local optima can be enhanced, thereby improving the performance of the trained machine learning model.

[0102] In step S13, the data in the model training dataset is first divided into a training set (80%) and a test set (20%). Then, the model is trained using a 10-fold cross-training method with machine learning algorithms such as Lightweight Gradient Boosting Machine (LightGBM), Extreme Gradient Boosting (XGBoost), Random Forest (RF), Support Vector Machine (SVM), and Extra-Trees (ET). During the training process, hyperparameters are optimized, for example, with F1 as the optimization objective, to obtain hyperparameter combinations for model training.

[0103] S14. Based on the model performance of multiple machine learning models obtained from training, the machine learning model with the best model performance is selected as the rock mass condition prediction model.

[0104] In one example, step S14 includes:

[0105] The first step is to sum the model performance metrics of the trained machine learning model to obtain the sum of the model performance metrics of the machine learning model. The model performance metrics include the weighted average of accuracy and recall (F1), accuracy (Acurracy), and the area under the receiver operating feature curve (AUC).

[0106] The second step is to use the machine learning model with the largest sum of the model performance indicators as the rock mass condition prediction model.

[0107] In this embodiment of the disclosure, model performance metrics are used to evaluate the model performance of the trained machine learning model. By determining the values ​​of F1, Acurracy, and AUC in the model performance metrics, and by determining the sum of F1, Acurracy, and AUC, the model performance of the multiple machine learning models finally trained can be evaluated, and the model with the best performance can be used for rock mass condition prediction.

[0108] S15. Obtain the raw tunneling data of the tunnel boring machine.

[0109] S16. Perform feature filtering and feature construction on the raw tunneling data of the tunnel boring machine to form a tunneling dataset.

[0110] See step S12.

[0111] S17. Input the tunneling dataset into the rock mass condition prediction model to predict the rock mass conditions during the tunneling process.

[0112] This disclosure provides a method for predicting rock mass conditions. It obtains raw tunneling data from a tunnel boring machine and the corresponding rock mass condition levels as raw data. Feature filtering and feature construction are performed on the raw data to form a model training dataset. The data in the model training dataset is more conducive to the performance of the trained machine learning model compared to the raw data. During model training, the Northern Eagle optimization algorithm, improved with a cross-cutting strategy, is used to optimize the hyperparameters, ensuring that the hyperparameter combination used in the machine learning model training is optimal, thereby further improving the model performance. Finally, by evaluating the model performance metrics, the model with the best performance is selected as the rock mass condition prediction model, ensuring the model's prediction accuracy.

[0113] Figure 2 This is a complete cycle diagram of TBM tunneling operation parameters provided in an embodiment of this disclosure.

[0114] Figure 3 A flowchart illustrating a raw data processing procedure provided in an embodiment of this disclosure. See also... Figure 3 , Figure 3The process of processing the raw data is shown, with the data sequence number indicating the processing step.

[0115] Figure 4 A flowchart illustrating model training as provided in an embodiment of this disclosure. See also... Figure 4 , Figure 4 The process of model training is shown in the figure.

[0116] Figure 5 This is a structural block diagram of a rock mass condition prediction device provided in an embodiment of this disclosure. See also... Figure 5 ,include:

[0117] The acquisition module 21 is used to acquire the original tunneling data of the TBM and the rock mass condition level corresponding to the original tunneling data as the original data; and to acquire the original tunneling data of the tunnel boring machine.

[0118] The data processing module 22 is used to perform feature filtering and feature construction based on the raw data to form a model training dataset; and to perform feature filtering and feature construction on the raw tunneling data of the tunnel boring machine to form a tunneling dataset.

[0119] The model training module 23 is used to train the model using multiple machine learning model training algorithms based on the training set in the model training dataset. During the model training process, the Northern Eagle optimization algorithm, which is improved through the cross-hatching strategy, is used to optimize the hyperparameters during the model training process. The test set is used to test the model performance of the multiple machine learning models trained, and the machine learning model with the best performance is used as the rock mass condition prediction model.

[0120] The prediction module 24 is used to input the tunneling dataset into the rock mass condition prediction model to predict the rock mass conditions during the tunneling process.

[0121] Optionally, the data processing module 22 is used for:

[0122] The Isolation Forest algorithm is used to remove outliers from the original data and fill in the null values ​​in the original data to form the first tunneling parameter dataset;

[0123] Feature filtering is performed on the first tunneling parameter dataset. The variance of a single feature and the correlation between features are calculated. Features with variances below the variance threshold or correlations above the correlation threshold are removed to obtain the second tunneling parameter dataset.

[0124] Based on the data in the first tunneling parameter dataset, feature construction is performed to obtain a tunneling performance feature dataset, which includes thrust penetration index, torque penetration index and specific penetration.

[0125] The second tunneling parameter dataset and the tunneling performance feature dataset are merged to obtain the model training dataset.

[0126] Optionally, the data processing module 22 is used for:

[0127] According to the formula Feature construction was performed to obtain the field thrust penetration index. The Indicates thrust, the Indicates the number of hobs, the Indicates penetration degree;

[0128] According to the formula Feature construction was performed to obtain the field torque penetration index. The Indicates the torque of the cutter head, the Indicates the diameter of the cutter head;

[0129] According to the formula Feature construction is performed to obtain the penetration degree. .

[0130] Optionally, the model training module 23 is used for:

[0131] Step 1, Initialize the Northern Goshawk population , The This represents the population matrix of the Northern Goshawk. Indicates the first The location of the northern goshawk, the aforementioned Indicates the first The first Northern Eagle The position of each dimension of the Northern Goshawk represents the value of a hyperparameter.

[0132] Step 2: Determine the value of a certain model performance index of the machine learning model corresponding to the Northern Goshawk based on the location of the Northern Goshawk, and use it as the fitness value; the model performance index includes any one of the following: the weighted average of accuracy and recall, accuracy, and the area under the receiver operating feature curve.

[0133] Step 3, use the formula , , To identify and attack prey; Indicates the first After the first Northern Goshawk identified and attacked its prey, the second... Wei's new position; Indicates the first Fitness value of a Northern Goshawk; Indicates the first The location of the prey of the northern eagle; Indicates the first The first of the prey of the northern goshawk The position of the dimension; Indicates the first The fitness value of the prey of the Northern Goshawk; Indicates the first Target location update during prey identification and attack operations by a Northern Goshawk; Indicates the first step in prey identification and attack operations. The fitness value of the target location of a Northern Goshawk is updated. for Random numbers within; It is a random number, either 1 or 2;

[0134] Step 4, use the formula , , They conducted chase and escape maneuvers; among them, Indicates the first After the Northern Goshawk completed its pursuit and escape maneuvers, the... Wei's new position; Indicates the first The target location is updated when a Northern Goshawk performs a chase and escape maneuver. Indicates the first The fitness value of the target's updated position when a Northern Goshawk performs a chase and escape maneuver; Indicates the current iteration number; Indicates the maximum number of iterations;

[0135] Step 5: Perform a horizontal cross operation between the position of the Northern Goshawk obtained in Step 2 and the position of the Northern Goshawk obtained in Step 4.

[0136] Step 6: Perform a vertical cross-operation on the position of the Northern Goshawk in Step 2;

[0137] Step 7: Compare the fitness values ​​of the Northern Goshawks before and after the longitudinal crossing in Step 6, and retain the Northern Goshawks with higher fitness values.

[0138] Step 8: Repeat steps 2 through 7 until the maximum number of iterations is met. At that time, the coordinates of the Northern Goshawk with the highest fitness value are output as the hyperparameter combination of the machine learning model.

[0139] Optionally, the model training module 23 is used for:

[0140] According to the formula , The location of the Northern Goshawk in step 2 The location of the Northern Goshawk obtained in step 4 The Dimensions intersect horizontally; among them, This indicates the first step in step 2. The first Northern Eagle Dimensional position; This indicates the result obtained in step 4. The first Northern Eagle Dimensional position; This indicates the first step in step 2. The offspring of the Northern Goshawk obtained by lateral crossbreeding dimension; This indicates the result obtained in step 4. The offspring of the Northern Goshawk obtained by lateral crossbreeding dimension; and for A random number between [a certain number of points].

[0141] Optionally, the model training module 23 is used for:

[0142] According to the formula The Northern Goshawk in step 2 The peacekeeping Dimensions are vertically intersected; among them, To bring the Northern Goshawk in step 2 The peacekeeping The offspring individuals obtained after vertical crossover of dimensions dimension.

[0143] Optionally, the model training module 23 is used for:

[0144] The weighted average of the accuracy and recall of the trained machine learning model, the area under the curve of the accuracy and receiver operating feature curve are summed to obtain the sum of the model performance metrics of the machine learning model.

[0145] The machine learning model that accumulates the highest sum of the model performance indicators is used as the rock mass condition prediction model.

[0146] Figure 6 This is a structural block diagram of an electronic device provided in an embodiment of this disclosure. See also... Figure 6 Electronic devices may include Figure 5 The aforementioned rock mass condition prediction device typically includes a processor 31 and a memory 32.

[0147] Processor 31 may include one or more processing cores, such as a 4-core processor, an 8-core processor, etc. Processor 31 may be implemented using at least one hardware form selected from DSP (Digital Signal Processing), FPGA (Field-Programmable Gate Array), and PLA (Programmable Logic Array). Processor 31 may also include a main processor and a coprocessor. The main processor is used to process data in the wake-up state, also known as a CPU (Central Processing Unit); the coprocessor is a low-power processor used to process data in the standby state. Memory 32 may include one or more computer-readable storage media, which may be non-transitory. Memory 32 may also include high-speed random access memory and non-volatile memory, such as one or more disk storage devices or flash memory devices. In some embodiments, the non-transitory computer-readable storage media in memory 32 is used to store at least one instruction, which is executed by processor 31 to implement the rock mass condition prediction method executed by an electronic device provided in the method embodiments of this application.

[0148] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for predicting rock mass conditions, characterized in that, include: Obtain the raw tunneling data of the tunnel boring machine and the corresponding rock mass condition level as the raw data; Based on the raw data, feature selection and feature construction are performed to form a model training dataset; Based on the model training dataset, multiple machine learning model training algorithms were used for model training. During the model training process, the Northern Eagle optimization algorithm, which was improved through a cross-hatching strategy, was used to optimize the hyperparameters during the model training process. Based on the model performance of multiple machine learning models obtained from training, the machine learning model with the best model performance is selected as the rock mass condition prediction model. Obtain the raw tunneling data from the tunnel boring machine; The raw tunneling data from the tunnel boring machine is subjected to feature filtering and feature construction to form a tunneling dataset; The tunneling dataset is input into the rock mass condition prediction model to predict the rock mass conditions during the tunneling process; The improved Northern Eagle optimization algorithm, refined through a cross-sectional strategy, optimizes hyperparameters during model training, including: Step 1, Initialize the Northern Goshawk population , The This represents the population matrix of the Northern Goshawk. Indicates the first The location of the northern goshawk, the aforementioned Indicates the first The first Northern Eagle The position of each dimension of the Northern Goshawk represents the value of a hyperparameter. Step 2: Determine the value of a certain model performance index of the machine learning model corresponding to the Northern Goshawk based on the location of the Northern Goshawk, and use it as the fitness value; the model performance index includes any one of the following: the weighted average of accuracy and recall, accuracy, and the area under the receiver operating feature curve. Step 3, use the formula , , To identify and attack prey; Indicates the first After the first Northern Goshawk identified and attacked its prey, the second... Wei's new position; Indicates the first Fitness value of a Northern Goshawk; Indicates the first The location of the prey of the northern eagle; Indicates the first The first of the prey of the northern goshawk The position of the dimension; Indicates the first The fitness value of the prey of the Northern Goshawk; Indicates the first Target location update during prey identification and attack operations by a Northern Goshawk; Indicates the first step in prey identification and attack operations. The fitness value of the target location of a Northern Goshawk is updated. for Random numbers within; It is a random number, either 1 or 2; Step 4, use the formula , , They conducted chase and escape maneuvers; among them, Indicates the first After the Northern Goshawk completed its pursuit and escape maneuvers, the... Wei's new position; Indicates the first The target location is updated when a Northern Goshawk performs a chase and escape maneuver. Indicates the first The fitness value of the target's updated position when a Northern Goshawk performs a chase and escape maneuver; Indicates the current iteration number; Indicates the maximum number of iterations; Step 5: Perform a horizontal cross operation between the position of the Northern Goshawk obtained in Step 2 and the position of the Northern Goshawk obtained in Step 4. Step 6: Perform a vertical cross-operation on the position of the Northern Goshawk in Step 2; Step 7: Compare the fitness values ​​of the Northern Goshawks before and after the longitudinal crossing in Step 6, and retain the Northern Goshawks with higher fitness values. Step 8: Repeat steps 2 through 7 until the maximum number of iterations is met. At that time, the coordinates of the Northern Goshawk with the highest fitness value are output as the hyperparameter combination of the machine learning model; Perform a horizontal cross operation between the location of the Northern Goshawk in step 2 and the location of the Northern Goshawk obtained in step 4, including: According to the formula , The location of the Northern Goshawk in step 2 The location of the Northern Goshawk obtained in step 4 The Dimensions intersect horizontally; among them, This indicates the first step in step 2. The first Northern Eagle Dimensional position; This indicates the result obtained in step 4. The first Northern Eagle Dimensional position; This indicates the first step in step 2. The offspring of the Northern Goshawk obtained by lateral crossbreeding dimension; This indicates the result obtained in step 4. The offspring of the Northern Goshawk obtained by lateral crossbreeding dimension; and for Random numbers between; Perform a vertical cross-operation on the position of the Northern Goshawk in step 2, including: According to the formula The Northern Goshawk in step 2 The peacekeeping Dimensions are vertically intersected; among them, To bring the Northern Goshawk in step 2 The peacekeeping The offspring individuals obtained after vertical crossover of dimensions dimension.

2. The rock mass condition prediction method according to claim 1, characterized in that, Based on the raw data, feature selection and feature construction are performed to form the model training dataset, including: The Isolation Forest algorithm is used to remove outliers from the original data and fill in the null values ​​in the original data to form the first tunneling parameter dataset; Feature filtering is performed on the first tunneling parameter dataset. The variance of a single feature and the correlation between features are calculated. Features with variances below the variance threshold or correlations above the correlation threshold are removed to obtain the second tunneling parameter dataset. Based on the data in the first tunneling parameter dataset, feature construction is performed to obtain a tunneling performance feature dataset, which includes thrust penetration index, torque penetration index and specific penetration. The second tunneling parameter dataset and the tunneling performance feature dataset are merged to obtain the model training dataset.

3. The rock mass condition prediction method according to claim 2, characterized in that, Based on the data in the first tunneling parameter dataset, feature construction is performed to obtain a tunneling performance feature dataset. This dataset includes thrust penetration index, torque penetration index, and specific penetration, comprising: According to the formula Feature construction was performed to obtain the field thrust penetration index. The Indicates thrust, the Indicates the number of hobs, the Indicates penetration degree; According to the formula Feature construction was performed to obtain the field torque penetration index. The Indicates the torque of the cutter head, the Indicates the diameter of the cutter head; According to the formula Feature construction is performed to obtain the penetration degree. .

4. The rock mass condition prediction method according to any one of claims 1 to 3, characterized in that, Based on the performance of multiple machine learning models trained, the best-performing machine learning model is selected as the rock mass condition prediction model, including: The weighted average of the accuracy and recall of the trained machine learning model, the area under the curve of the accuracy and receiver operating feature curve are summed to obtain the sum of the model performance metrics of the machine learning model. The machine learning model that accumulates the highest sum of the model performance indicators is used as the rock mass condition prediction model.

5. A rock mass condition prediction device, characterized in that, include: The acquisition module acquires the raw tunneling data of the tunnel boring machine and the corresponding rock mass condition level as the raw data; Obtain the raw tunneling data from the tunnel boring machine; The data processing module is used to perform feature filtering and feature construction based on the raw data to form a model training dataset; it also performs feature filtering and feature construction on the raw tunneling data of the tunnel boring machine to form a tunneling dataset. The model training module is used to train models using multiple machine learning model training algorithms based on the model training dataset. During the model training process, the Northern Eagle optimization algorithm, which is improved through a cross-cutting strategy, is used to optimize the hyperparameters of the model training process. Based on the model performance of the multiple machine learning models obtained from the training, the machine learning model with the best performance is selected as the rock mass condition prediction model. The prediction module is used to input the tunneling dataset into the rock mass condition prediction model to predict the rock mass conditions during the tunneling process; The improved Northern Eagle optimization algorithm, refined through a cross-sectional strategy, optimizes hyperparameters during model training, including: Step 1, Initialize the Northern Goshawk population , The This represents the population matrix of the Northern Goshawk. Indicates the first The location of the northern goshawk, the aforementioned Indicates the first The first Northern Eagle The position of each dimension of the Northern Goshawk represents the value of a hyperparameter. Step 2: Determine the value of a certain model performance index of the machine learning model corresponding to the Northern Goshawk based on the location of the Northern Goshawk, and use it as the fitness value; the model performance index includes any one of the following: the weighted average of accuracy and recall, accuracy, and the area under the receiver operating feature curve. Step 3, use the formula , , To identify and attack prey; Indicates the first After the first Northern Goshawk identified and attacked its prey, the second... Wei's new position; Indicates the first Fitness value of a Northern Goshawk; Indicates the first The location of the prey of the northern eagle; Indicates the first The first of the prey of the northern goshawk The position of the dimension; Indicates the first The fitness value of the prey of the Northern Goshawk; Indicates the first Target location update during prey identification and attack operations by a Northern Goshawk; Indicates the first step in prey identification and attack operations. The fitness value of the target location of a Northern Goshawk is updated. for Random numbers within; It is a random number, either 1 or 2; Step 4, use the formula , , They conducted chase and escape maneuvers; among them, Indicates the first After the Northern Goshawk completed its pursuit and escape maneuvers, the... Wei's new position; Indicates the first The target location is updated when a Northern Goshawk performs a chase and escape maneuver. Indicates the first The fitness value of the target's updated position when a Northern Goshawk performs a chase and escape maneuver; Indicates the current iteration number; Indicates the maximum number of iterations; Step 5: Perform a horizontal cross operation between the position of the Northern Goshawk obtained in Step 2 and the position of the Northern Goshawk obtained in Step 4. Step 6: Perform a vertical cross-operation on the position of the Northern Goshawk in Step 2; Step 7: Compare the fitness values ​​of the Northern Goshawks before and after the longitudinal crossing in Step 6, and retain the Northern Goshawks with higher fitness values. Step 8: Repeat steps 2 through 7 until the maximum number of iterations is met. At that time, the coordinates of the Northern Goshawk with the highest fitness value are output as the hyperparameter combination of the machine learning model; Perform a horizontal cross operation between the location of the Northern Goshawk in step 2 and the location of the Northern Goshawk obtained in step 4, including: According to the formula , The location of the Northern Goshawk in step 2 The location of the Northern Goshawk obtained in step 4 The Dimensions intersect horizontally; among them, This indicates the first step in step 2. The first Northern Eagle Dimensional position; This indicates the result obtained in step 4. The first Northern Eagle Dimensional position; This indicates the first step in step 2. The offspring of the Northern Goshawk obtained by lateral crossbreeding dimension; This indicates the result obtained in step 4. The offspring of the Northern Goshawk obtained by lateral crossbreeding dimension; and for Random numbers between; Perform a vertical cross-operation on the position of the Northern Goshawk in step 2, including: According to the formula The Northern Goshawk in step 2 The peacekeeping Dimensions are vertically intersected; among them, To bring the Northern Goshawk in step 2 The peacekeeping The offspring individuals obtained after vertical crossover of dimensions dimension.

6. An electronic device, characterized in that, Includes the rock mass condition prediction device as described in claim 5.

7. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores at least one piece of program code, which is executed by a processor to implement the rock mass condition prediction method as described in any one of claims 1 to 4.

Images