Construction method and device of tunnel extrusion deformation prediction model

Abstract

The invention provides a construction method and device of a tunnel extrusion deformation prediction model. Comprising the steps of obtaining case data of tunnel extrusion deformation, and generating an initial data set; performing data filling on the initial data set by adopting a plurality of data filling methods to generate a plurality of complete data sets; training a plurality of first machine learning models according to the complete data set, and determining a model training data set according to model performance indexes of the plurality of first machine learning models; training a plurality of second machine learning models according to a training set in the model training data set, and optimizing hyper-parameters in the training process of the second machine learning models by adopting an improved black wing plinergic algorithm in the model training process; and determining a tunnel extrusion deformation prediction model according to the model performance indexes of the plurality of second machine learning models obtained by training. The model performance of the tunnel extrusion deformation prediction model obtained through training can be improved.

Description

Technical Field

[0001] The invention belongs to the field of tunnel engineering, and in particular relates to a method and a device for constructing a tunnel extrusion deformation prediction model. Background Art

[0002] Due to the complexity of geological conditions, large deformation disasters often occur during tunnel construction. In previous engineering examples and projects, due to the limitations of construction sites and underground engineering conditions, most of them could not directly obtain complete geotechnical mechanical characteristics, resulting in the lack of complete data feature records in cases. The existing method for constructing tunnel extrusion deformation prediction models cannot form a more accurate tunnel extrusion deformation prediction model, and it is difficult to provide experience and guidance for whether large deformation disasters will occur in existing unexcavated tunnels and projects that have not yet been built. Summary of the invention

[0003] The present invention provides a method and device for constructing a tunnel extrusion deformation prediction model, which can improve the model performance of the tunnel extrusion deformation prediction model.

[0004] On the one hand, a method for constructing a tunnel extrusion deformation prediction model is provided, comprising: Obtain case data of tunnel extrusion deformation and generate an initial data set; Using multiple data filling methods to fill the initial data set, and generating multiple complete data sets; Training a plurality of first machine learning models according to the complete data set, and determining a model training data set according to model performance indicators of the plurality of first machine learning models; Training multiple second machine learning models according to the training sets in the model training data set, and using an improved black kite algorithm to optimize hyperparameters in the training process of the second machine learning models during the model training process; The tunnel extrusion deformation prediction model is determined according to the model performance indicators of multiple second machine learning models obtained through training.

[0005] Optionally, multiple data filling methods are used to fill the initial data set to generate multiple complete data sets, including: The initial data set was filled using mean imputation, median imputation, mode imputation, multiple imputation and KNN imputation methods to obtain multiple complete data sets.

[0006] Optionally, training multiple first machine learning models according to the complete data set, and determining the model training data set according to the model performance indicators of the multiple first machine learning models, includes: Based on each complete data set, multiple machine learning models are trained using different machine learning algorithms; Calculate the cumulative sum of model performance indicators of the machine learning model, where the model performance indicators include any one of the weighted average of accuracy and recall, accuracy, Matthews correlation coefficient, and Kappa coefficient; A machine learning model with the largest cumulative sum of model performance indicators is determined, where the complete data set used by the machine learning model with the largest cumulative sum of model performance indicators is used as the model training data set.

[0007] Optionally, multiple second machine learning models are trained according to the training sets in the model training data set, and during the model training process, an improved black kite algorithm is used to optimize hyperparameters in the second machine learning model training process, including: Taking the weighted average of accuracy and recall as the iterative optimization target, the improved black kite algorithm is used to optimize the hyperparameters in the training process of multiple machine learning model training algorithms; the improved black kite algorithm is obtained by using Logistic mapping to improve the initial population position of the black kite algorithm, and the initial population position of the improved black kite algorithm is ; According to the training set in the model training data set, training is performed using a ten-fold cross-validation training method and multiple optimized machine learning model training algorithms to obtain multiple second machine learning models.

[0008] Optionally, determining a tunnel extrusion deformation prediction model according to model performance indicators of multiple second machine learning models obtained through training includes: According to the second machine learning model, a model performance indicator of the second machine learning model training set is obtained, wherein the model performance indicator includes any one of a weighted average of accuracy and recall, accuracy, Matthews correlation coefficient, and Kappa coefficient; Testing the second machine learning model according to the test set to obtain a model performance indicator of the second machine learning model test set; The model performance indicators of the second machine learning model training set and test set are accumulated to obtain the cumulative sum of the model performance indicators of the second machine learning model. The second machine learning model with the largest cumulative sum of model performance indicators is the tunnel extrusion deformation prediction model.

[0009] On the other hand, a device for constructing a tunnel extrusion deformation prediction model is provided, characterized in that it includes: An acquisition module is used to acquire case data of tunnel extrusion deformation and generate an initial data set; A data filling module is used to fill the initial data set with data using multiple data filling methods to generate multiple complete data sets; A model training module is used to train multiple first machine learning models based on a complete data set, and determine a model training data set based on the model performance indicators of the multiple first machine learning models; train multiple second machine learning models based on the training sets in the model training data set, and use an improved black kite algorithm to optimize the hyperparameters in the second machine learning model training process during the model training process; and determine a tunnel extrusion deformation prediction model based on the model performance indicators of the multiple second machine learning models obtained through training.

[0010] Optionally, the model training module is used to: Based on each complete data set, multiple machine learning models are trained using different machine learning algorithms; Calculate the cumulative sum of model performance indicators of the machine learning model, where the model performance indicators include any one of the weighted average of accuracy and recall, accuracy, Matthews correlation coefficient, and Kappa coefficient; A machine learning model with the largest cumulative sum of model performance indicators is determined, where the complete data set used by the machine learning model with the largest cumulative sum of model performance indicators is used as the model training data set.

[0011] Optionally, the model training module is used to: Taking the weighted average of accuracy and recall as the iterative optimization target, the improved black kite algorithm is used to optimize the hyperparameters in the training process of multiple machine learning model training algorithms; the improved black kite algorithm is obtained by using Logistic mapping to improve the initial population position of the black kite algorithm, and the initial population position of the improved black kite algorithm is ; According to the training set in the model training data set, training is performed using a ten-fold cross-validation training method and multiple optimized machine learning model training algorithms to obtain multiple second machine learning models.

[0012] On the other hand, an electronic device is provided, comprising a device for constructing a tunnel extrusion deformation prediction model as described in any one of the above items.

[0013] On the other hand, a computer-readable storage medium is provided, in which at least one program code is stored, and the program code is executed by a processor to implement the method for constructing a tunnel extrusion deformation prediction model as described in any one of the above items.

[0014] The technical solution provided by the embodiments of the present disclosure brings the following beneficial effects: In an embodiment of the present disclosure, a method for constructing a tunnel extrusion deformation prediction model is provided. By acquiring case data, multiple data filling methods are used to fill the case data to obtain multiple complete data sets, thereby ensuring the completeness of the case data. After obtaining the complete data set, different algorithms are used to train the machine learning model according to the complete data set, and then the model performance indicators of all machine learning models are evaluated to determine the optimal data, that is, the data in the model training data set, to ensure the model performance of the tunnel extrusion deformation prediction model obtained by subsequent training. After determining the model training data set, multiple second machine learning models are trained according to the model training data set, and the improved black-winged kite algorithm is used to optimize the hyperparameters in the model training process to ensure the model performance of the second machine learning model obtained by training. After obtaining multiple second machine learning models, the model performance indicators of all second machine learning models are counted, and the second machine learning model with the best performance is the tunnel extrusion deformation prediction model. Through the method for constructing a tunnel extrusion deformation prediction model provided by the present invention, it can be ensured that the model performance indicators of the tunnel extrusion deformation prediction model finally obtained by training are good. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0016] Figure 1 A method flow chart of a method for constructing a tunnel extrusion deformation prediction model provided by an embodiment of the present disclosure; Figure 2 A flowchart of filling an initial data set provided by an embodiment of the present disclosure; Figure 3 A flowchart of an improved black kite algorithm provided by an embodiment of the present disclosure; Figure 4 A flowchart of a second machine learning model training process provided in an embodiment of the present disclosure; Figure 5 A structural block diagram of a device for constructing a tunnel extrusion deformation prediction model provided by an embodiment of the present disclosure; Figure 6 A structural block diagram of an electronic device provided in an embodiment of the present disclosure.

[0017] The reference numerals are as follows: 21: acquisition module; 22: data filling module; 23: model training module; 31: processor; 32: memory. DETAILED DESCRIPTION

[0018] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0019] The variables usually used to predict large extrusion deformation disasters include: tunnel depth (H), tunnel diameter (D), support stiffness (K), rock mass quality index (Q), and strength stress ratio (SSR). However, in the existing engineering records, tunnel deformation can be easily obtained based on on-site survey measurements, and geotechnical characteristics such as Q, K and SSR cannot be fully recorded. Therefore, it is not possible to analyze extrusion deformation disasters based on incomplete data sets collected directly. It is necessary to first fill in the data records, and then establish an extrusion deformation prediction model based on the complete data set. Hoek proposed the large extrusion deformation classification principle based on the tunnel extrusion deformation: let ε (%) be the relative deformation of the tunnel, ε>10 is extremely severe extrusion deformation, 10>ε>5 is severe extrusion deformation, 5>ε>2.5 is moderate extrusion deformation, 2.5>ε>1 is slight extrusion deformation, and ε<1 is no extrusion deformation. According to the variable characteristics and the large extrusion deformation classification principle, a large deformation grade prediction model is finally established.

[0020] Figure 1 A method flow chart of a method for constructing a tunnel extrusion deformation prediction model provided by an embodiment of the present disclosure. Figure 1 ,include: S11. Obtain case data of tunnel extrusion deformation and generate an initial data set.

[0021] In step S11, historical tunnel engineering extrusion deformation case data from various countries and regions are collected, including tunnel burial depth (H), tunnel diameter (D), support stiffness (K), rock mass quality index (Q), strength stress ratio (SSR) and tunnel extrusion deformation grade (Rank). The missing characteristic values ​​in the project are temporarily marked as Null, forming an incomplete data set of tunnel engineering extrusion deformation disasters.

[0022] S12. Use multiple data filling methods to fill the initial data set to generate multiple complete data sets.

[0023] In one example, step S12 includes: The initial data set was filled using mean imputation, median imputation, mode imputation, multiple imputation and KNN imputation methods to obtain multiple complete data sets.

[0024] In the embodiments of the present disclosure, different data filling methods are used to fill in different data in the complete data set. By using multiple data filling methods to fill in the data in the initial data set, it can be ensured that the obtained complete data set is relatively rich, which is conducive to determining the data set that is most conducive to model training in the subsequent steps.

[0025] S13. Train multiple first machine learning models based on the complete data set, and determine the model training data set based on the model performance indicators of the multiple first machine learning models.

[0026] In one example, step S13 includes: The first step is to use different machine learning algorithms to train multiple machine learning models based on each complete data set.

[0027] In the first step, multiple machine learning algorithms are support vector machine (SVM) algorithm, random forest (RF) algorithm, logistic regression (LR) algorithm, adaptive boosting (AdaBoost) algorithm, and extreme boosting (XGBoost) algorithm.

[0028] Of course, the present invention is not limited to the above algorithm, which is just an example provided by the present invention.

[0029] The second step is to calculate the cumulative sum of model performance indicators of the machine learning model, where the model performance indicators include any one of the weighted average of accuracy and recall (F1), accuracy (ACC), Matthews correlation coefficient (MCC), and Kappa coefficient.

[0030] In the embodiments of the present disclosure, the model performance index may not be limited to the above-mentioned model performance index.

[0031] The third step is to determine the machine learning model with the largest cumulative sum of model performance indicators, where the complete data set used by the machine learning model with the largest cumulative sum of model performance indicators is the model training data set.

[0032] In the disclosed embodiment, each complete data set uses multiple machine learning algorithms to perform machine learning model training. After all complete data sets are trained, multiple first machine learning models are obtained. The model performance of the first machine learning model can be determined by evaluating the cumulative model performance indicators of the first machine learning models. The data set used by the first machine learning model with the best model performance is most conducive to machine learning model training.

[0033] S14. Train multiple second machine learning models according to the training sets in the model training data set, and use the improved black kite algorithm to optimize the hyperparameters in the second machine learning model training process during the model training process.

[0034] In one example, step S14 includes: In the first step, taking F1 as the iterative optimization target, the improved black kite algorithm is used to optimize the hyperparameters in the training process of multiple machine learning model training algorithms; the improved black kite algorithm is obtained by using Logistic mapping to improve the initial population position of the black kite algorithm, and the initial population position of the improved black kite algorithm is .

[0035] In the existing black kite algorithm, the initial position is ,in, , Respectively represent The upper and lower bounds of the black-winged kite individual. is a random number between [0,1].

[0036] In the embodiment of the present disclosure, by introducing Logistic mapping to modify the black-winged kite (BKA) to form an improved black-winged kite (LBKA) algorithm, the richness of the algorithm population individuals can be increased, and the possibility of escaping from the local optimum can be increased.

[0037] In the second step, according to the training set in the model training data set, a ten-fold cross-validation training method and multiple optimized machine learning model training algorithms are used for training to obtain multiple second machine learning models.

[0038] In the second step, the multiple second machine learning models obtained through training are LBKA-SVM model, LBKA-RF model, LBKA-LR model, LBKA-AdaBoost model, and LBKA-XGBoost model, which are obtained by optimizing the hyperparameters of SVM, RF, LR, AdaBoost, and XGBoost by LBKA respectively.

[0039] In the disclosed embodiment, the initial population position of the black kite algorithm is improved by logistic mapping to form an improved black kite algorithm (denoted as LBKA). By modifying the initial population position of the black kite algorithm by logistic mapping, and then performing population attack and population migration, the richness of population iteration and the operation convergence speed of the improved black kite algorithm can be increased. By optimizing the hyperparameters in the training process of the machine learning model by the improved black kite algorithm, the model performance of the machine learning obtained by training can be improved.

[0040] In the disclosed embodiment, the attack behavior of the improved black kite algorithm (LBKA) is as follows: , ; in, and Indicates A black kite The , The position of the iteration step; is a random number in [0,1], =0.9; is the total number of iterations; is the number of iterations completed so far.

[0041] The migration behavior of the improved black kite algorithm (LBKA) is as follows: , ; in, For the current The iteration Leading scorer for the Black-winged Kite; For the In the iteration, any (i.e., a certain) black-winged kite The fitness value of a random position in the dimension; is the number of any (that is, any) black-winged kite in the tth iteration The fitness value of the dimensional position; It is a Cauchy mutation.

[0042] In the embodiment of the present disclosure, the data in the model training data set is divided into a training set and a test set, and the training set and the test set are obtained by randomly selecting data in the model training data set.

[0043] In one example, 80% of the data in the model training data set is randomly selected as the training set, and the remaining 20% ​​is used as the test set.

[0044] Of course, the above is only an example of selecting training set and test set data, and the present disclosure is not limited to this.

[0045] S15. Determine a tunnel extrusion deformation prediction model based on the model performance indicators of multiple second machine learning models obtained through training.

[0046] In one example, step S15 includes: The first step is to obtain model performance indicators of the second machine learning model training set according to the second machine learning model, where the model performance indicators include ACC, F1, Kappa, and MCC.

[0047] The second step is to test the second machine learning model based on the test set to obtain the model performance index of the second machine learning model test set.

[0048] The third step is to accumulate the model performance indicators of the second machine learning model training set and test set to obtain the cumulative sum of the model performance indicators of the second machine learning model. The second machine learning model with the largest cumulative sum of model performance indicators is the tunnel extrusion deformation prediction model.

[0049] In the disclosed embodiment, during the model training process, the improved black kite algorithm is used to optimize the hyperparameters in the model training process, so the model performance of the second machine learning model finally obtained by training will be changed to a certain extent compared with the first machine learning model. By evaluating the model performance of the multiple second machine learning models finally obtained, the second machine learning model with the best model performance is the tunnel extrusion deformation prediction model.

[0050] In an embodiment of the present disclosure, a method for constructing a tunnel extrusion deformation prediction model is provided. By acquiring case data, multiple data filling methods are used to fill the case data to obtain multiple complete data sets, thereby ensuring the completeness of the case data. After obtaining the complete data set, different algorithms are used to train the machine learning model according to the complete data set, and then the model performance indicators of all machine learning models are evaluated to determine the optimal data, that is, the data in the model training data set, to ensure the model performance of the tunnel extrusion deformation prediction model obtained by subsequent training. After determining the model training data set, multiple second machine learning models are trained according to the model training data set, and the improved black-winged kite algorithm is used to optimize the hyperparameters in the model training process to ensure the model performance of the second machine learning model obtained by training. After obtaining multiple second machine learning models, the model performance indicators of all second machine learning models are counted, and the second machine learning model with the best performance is the tunnel extrusion deformation prediction model. Through the method for constructing a tunnel extrusion deformation prediction model provided by the present invention, it can be ensured that the model performance indicators of the tunnel extrusion deformation prediction model finally obtained by training are good.

[0051] Figure 2 A flowchart of filling an initial data set provided by an embodiment of the present disclosure. Figure 2 The initial data set includes case data of various levels of extrusion deformation. The average imputation method, median imputation method, mode imputation method, multiple interpolation method and KNN interpolation method are used to impute the initial data set, thereby obtaining multiple complete data sets (stored in the complete database). The number of complete data sets corresponds to the number of data imputation methods used.

[0052] After the filling is completed, the model training algorithm in the classifier group is used for training to obtain multiple first machine learning models. The cumulative sum of the model performance of the first machine learning models is evaluated to determine the final database (the final database stores the model training data set).

[0053] Figure 3 A flowchart of an improved black kite algorithm provided by an embodiment of the present disclosure. Figure 3 Compared with the Black Kite Algorithm (BKA), the improved Black Kite Algorithm (LBKA) uses a modified Logistic mapping to modify the initial population position of the Black Kite Algorithm, thereby improving the richness of population iteration and the convergence speed of operation.

[0054] Figure 4 A flowchart of a second machine learning model training process provided by an embodiment of the present disclosure. Figure 4 , the second machine learning model is based on Figure 2 The complete data set (i.e., the model training data set) determined in the training is used for training, and the improved black kite algorithm (LBKA) is used to optimize the hyperparameters in the machine learning model training process ( Figure 4 After the second machine learning model is trained, a comprehensive evaluation of the model performance is performed, and the final extrusion deformation grade prediction model (i.e., tunnel extrusion deformation prediction model) is obtained based on the evaluation results.

[0055] Figure 5 This is a structural block diagram of a device for constructing a tunnel extrusion deformation prediction model provided by an embodiment of the present disclosure. Figure 5 ,include: The acquisition module 21 is used to acquire case data of tunnel extrusion deformation and generate an initial data set.

[0056] The data filling module 22 is used to fill the initial data set with data using multiple data filling methods to generate multiple complete data sets.

[0057] The model training module 23 is used to train multiple first machine learning models based on the complete data set, and determine the model training data set based on the model performance indicators of the multiple first machine learning models; train multiple second machine learning models based on the training sets in the model training data set, and use the improved black kite algorithm to optimize the hyperparameters in the second machine learning model training process during the model training process; determine the tunnel extrusion deformation prediction model based on the model performance indicators of the multiple second machine learning models obtained through training.

[0058] Optionally, the data filling module 22 is used to: The initial data set was filled using mean imputation, median imputation, mode imputation, multiple imputation and KNN imputation methods to obtain multiple complete data sets.

[0059] Optionally, the model training module 23 is used to: Based on each complete data set, multiple machine learning models are trained using different machine learning algorithms; Calculate the cumulative sum of model performance indicators of the machine learning model, where the model performance indicators include ACC, F1, Kappa, and MCC; A machine learning model with the largest cumulative sum of model performance indicators is determined, where the complete data set used by the machine learning model with the largest cumulative sum of model performance indicators is used as the model training data set.

[0060] Optionally, the model training module 23 is used to: Taking F1 as the iterative optimization target, the improved black kite algorithm is used to optimize the hyperparameters in the training process of multiple machine learning model training algorithms; the improved black kite algorithm is obtained by using Logistic mapping to improve the initial population position of the black kite algorithm, and the initial population position of the improved black kite algorithm is ; According to the training set in the model training data set, training is performed using a ten-fold cross-validation training method and multiple optimized machine learning model training algorithms to obtain multiple second machine learning models.

[0061] Optionally, the model training module 23 is used to: According to the second machine learning model, a model performance index of the second machine learning model training set is obtained, wherein the model performance index includes ACC, F1, Kappa, and MCC; Testing the second machine learning model according to the test set to obtain a model performance indicator of the second machine learning model test set; The model performance indicators of the second machine learning model training set and test set are accumulated to obtain the cumulative sum of the model performance indicators of the second machine learning model. The second machine learning model with the largest cumulative sum of model performance indicators is the tunnel extrusion deformation prediction model.

[0062] Figure 6 This is a structural block diagram of an electronic device provided by an embodiment of the present disclosure. Figure 6 , electronic devices may include Figure 5 The device for constructing the tunnel extrusion deformation prediction model. Generally, the electronic device includes: a processor 31 and a memory 32.

[0063] The processor 31 may include one or more processing cores, such as a 4-core processor, an 8-core processor, etc. The processor 31 may be implemented in at least one hardware form of DSP (Digital Signal Processing), FPGA (Field-Programmable Gate Array), and PLA (Programmable Logic Array). The processor 31 may also include a main processor and a coprocessor. The main processor is a processor for processing data in the awake state, also known as a CPU (Central Processing Unit); the coprocessor is a low-power processor for processing data in the standby state.

[0064] The memory 32 may include one or more computer-readable storage media, which may be non-transitory. The memory 32 may also include a high-speed random access memory, and a non-volatile memory, such as one or more disk storage devices, flash memory storage devices. In some embodiments, the non-transitory computer-readable storage medium in the memory 32 is used to store at least one instruction, which is used to be executed by the processor 31 to implement the method for constructing a tunnel extrusion deformation prediction model performed by an electronic device provided in the method embodiment of the present application.

[0065] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for constructing a tunnel extrusion deformation prediction model, characterized in that: include: Obtain case data of tunnel extrusion deformation and generate an initial data set; Using multiple data filling methods to fill the initial data set, and generating multiple complete data sets; Training a plurality of first machine learning models according to the complete data set, and determining a model training data set according to model performance indicators of the plurality of first machine learning models; Training multiple second machine learning models according to the model training data set, and using an improved black kite algorithm to optimize hyperparameters in the training process of the second machine learning models during the model training process; The tunnel extrusion deformation prediction model is determined according to the model performance indicators of multiple second machine learning models obtained through training.

2. The method for constructing a tunnel extrusion deformation prediction model according to claim 1, characterized in that: Multiple data filling methods are used to fill the initial data set to generate multiple complete data sets, including: The initial data set was filled using mean imputation, median imputation, mode imputation, multiple imputation and KNN imputation methods to obtain multiple complete data sets.

3. The method for constructing a tunnel extrusion deformation prediction model according to claim 1, characterized in that: Training multiple first machine learning models according to the complete data set, and determining the model training data set according to the model performance indicators of the multiple first machine learning models, including: Based on each complete data set, multiple machine learning models are trained using different machine learning algorithms; Calculate the cumulative sum of model performance indicators of the machine learning model, where the model performance indicators include any one of the weighted average of accuracy and recall, accuracy, Matthews correlation coefficient, and Kappa coefficient; A machine learning model with the largest cumulative sum of model performance indicators is determined, where the complete data set used by the machine learning model with the largest cumulative sum of model performance indicators is used as the model training data set.

4. The method for constructing a tunnel extrusion deformation prediction model according to claim 1, characterized in that: Training multiple second machine learning models according to the model training data set, and using the improved black kite algorithm to optimize the hyper parameters in the training process of the second machine learning model during the model training process, including: Taking the weighted average of accuracy and recall as the iterative optimization target, the improved black kite algorithm is used to optimize the hyperparameters in the training process of multiple machine learning model training algorithms; the improved black kite algorithm is obtained by using Logistic mapping to improve the initial population position of the black kite algorithm, and the initial population position of the improved black kite algorithm is ; According to the training set in the model training data set, training is performed using a ten-fold cross-validation training method and multiple optimized machine learning model training algorithms to obtain multiple second machine learning models.

5. The method for constructing a tunnel extrusion deformation prediction model according to claim 1, characterized in that: The tunnel extrusion deformation prediction model is determined according to the model performance indicators of the multiple second machine learning models obtained through training, including: According to the second machine learning model, a model performance indicator of the second machine learning model training set is obtained, wherein the model performance indicator includes any one of a weighted average of accuracy and recall, accuracy, Matthews correlation coefficient, and Kappa coefficient; Testing the second machine learning model according to the test set to obtain a model performance indicator of the second machine learning model test set; The model performance indicators of the second machine learning model training set and test set are accumulated to obtain the cumulative sum of the model performance indicators of the second machine learning model. The second machine learning model with the largest cumulative sum of model performance indicators is the tunnel extrusion deformation prediction model.

6. A device for constructing a tunnel extrusion deformation prediction model, characterized in that: include: An acquisition module is used to acquire case data of tunnel extrusion deformation and generate an initial data set; A data filling module is used to fill the initial data set with data using multiple data filling methods to generate multiple complete data sets; A model training module, used to train multiple first machine learning models based on the complete data set, and determine the model training data set based on the model performance indicators of the multiple first machine learning models; Multiple second machine learning models are trained according to the training sets in the model training data set. During the model training process, the improved black kite algorithm is used to optimize the hyperparameters in the second machine learning model training process; and the tunnel extrusion deformation prediction model is determined according to the model performance indicators of the multiple second machine learning models obtained through training.

7. The device for constructing a tunnel extrusion deformation prediction model according to claim 6, characterized in that: The model training module is used to: Based on each complete data set, multiple machine learning models are trained using different machine learning algorithms; Calculate the cumulative sum of model performance indicators of the machine learning model, where the model performance indicators include any one of the weighted average of accuracy and recall, accuracy, Matthews correlation coefficient, and Kappa coefficient; A machine learning model with the largest cumulative sum of model performance indicators is determined, where the complete data set used by the machine learning model with the largest cumulative sum of model performance indicators is used as the model training data set.

8. The device for constructing a tunnel extrusion deformation prediction model according to claim 6, characterized in that: The model training module is used to: Taking the weighted average of accuracy and recall as the iterative optimization target, the improved black kite algorithm is used to optimize the hyperparameters in the training process of multiple machine learning model training algorithms; the improved black kite algorithm is obtained by using Logistic mapping to improve the initial population position of the black kite algorithm, and the initial population position of the improved black kite algorithm is ; According to the training set in the model training data set, training is performed using a ten-fold cross-validation training method and multiple optimized machine learning model training algorithms to obtain multiple second machine learning models.

9. An electronic device, characterized in that: A device for constructing a tunnel extrusion deformation prediction model comprising the device described in any one of claims 6 to 8.

10. A computer-readable storage medium, characterized in that: At least one program code is stored in the computer-readable storage medium, and the program code is executed by a processor to implement the method for constructing a tunnel extrusion deformation prediction model as described in any one of claims 1 to 5.

Images