Tunnel overexcavation risk assessment method and system based on machine learning

Abstract

The invention relates to the field of tunnel engineering, and provides a tunnel overexcavation risk assessment method and system based on machine learning. The method comprises the following steps: training a prediction model based on acquired historical blasting design parameters, geological characteristics, an over-fault, a tunnel geometric shape and corresponding historical over-excavation values; in the training process, the optimal hyper-parameter of the prediction model is obtained by adopting a quilt group optimization algorithm, and the optimized prediction model is obtained; and obtaining blasting design parameters, geological characteristics, an over-fault and a tunnel geometrical shape, and adopting the optimized prediction model to obtain an over-excavation value so as to evaluate a tunnel over-excavation risk level.

Description

Technical Field

[0001] The present invention relates to the field of tunnel engineering, and in particular to a tunnel over-excavation risk assessment method and system based on machine learning. Background Art

[0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art.

[0003] Overbreak is a detrimental phenomenon caused by tunnel blasting that can increase the volume of rock expelled from the tunnel excavation and, therefore, the volume of additional shotcrete. In addition, it reduces the stability of the surrounding rock and causes local stress concentrations, which will lead to increased time and cost in the construction schedule. Due to the large amount of rock mass that will fall off and disintegrate during tunnel blasting, overbreak (OB) inevitably occurs and is prevalent with tunnel blasting operations, which leads to irreversible excavation damage zones. Therefore, it is very important to establish a model that can accurately predict the amount of overbreak caused by tunnel blasting.

[0004] At present, different scholars have conducted some research on OB caused by tunnel blasting, including empirical, statistical, numerical and artificial intelligence methods. The empirical method performs well in some specific engineering cases, but the factors considered by these equations are very simple and do not take into account all the influences and basic considerations about OB. Regarding statistical models, some scholars have tried to propose solutions to predict tunnel blasting-induced OB. However, it is overly dependent on the data that constitute the statistical results. Another type of technology used to solve the OB problem is numerical models. Numerical methods can simulate the generation process of OB very well, and even the emergence of OB can be observed intuitively, but the problem is that numerical methods are difficult to apply to high-dimensional problems. The OB phenomenon is not only related to one or two factors; accurate simulation requires consideration of more factors. In summary, although the above techniques (empirical, statistical and numerical) can achieve average level accuracy in predicting / simulating OB values, these methods cannot fully characterize the complex nonlinear relationship between OB and its influencing factors due to the increase in dimensionality and data volume. Summary of the invention

[0005] In order to solve the technical problems existing in the above-mentioned background technology, the present invention provides a tunnel over-excavation risk assessment method and system based on machine learning. The RF-TSA model designed by the present invention can provide more accurate over-excavation value and improve the accuracy of tunnel over-excavation risk assessment.

[0006] In order to achieve the above object, the present invention adopts the following technical solution:

[0007] A first aspect of the present invention provides a tunnel over-excavation risk assessment method based on machine learning.

[0008] A tunnel over-excavation risk assessment method based on machine learning, comprising:

[0009] Based on the historical blasting design parameters, geological characteristics, fault crossings, tunnel geometry, and corresponding historical overbreak values, the prediction model is trained. During the training process, the quilt group optimization algorithm is used to obtain the optimal hyperparameters of the prediction model and obtain the optimized prediction model.

[0010] The blasting design parameters, geological characteristics, fault crossings and tunnel geometry are obtained, and the over-excavation value is obtained using the optimized prediction model to evaluate the tunnel over-excavation risk level.

[0011] Furthermore, the process of using the quilt swarm optimization algorithm to obtain the optimal hyperparameters of the prediction model includes: initializing the population size and necessary parameters of the quilt swarm optimization algorithm; calculating the fitness value of the population and evaluating its fitness, saving the optimal fitness value and then improving the hyperparameters; updating the fitness value satisfied by the hyperparameters according to the results of each iteration to find the ideal hyperparameters; when the optimal fitness value no longer changes and the maximum number of iterations is reached, the optimal hyperparameters of the prediction model are obtained.

[0012] Furthermore, the prediction model adopts the RF model.

[0013] Furthermore, the optimal hyperparameters of the prediction model include: the number of decision trees in the forest, the maximum depth of the decision tree, and the minimum number of samples required to split an internal node.

[0014] Furthermore, the blasting design parameters include the number of blasting holes, hole depth, total charge, blasting length and unit consumption of explosives.

[0015] Furthermore, the cross-fault and tunnel geometries are measured by terrestrial laser scanning technology.

[0016] Furthermore, after obtaining the historical blasting design parameters, the method further includes: preprocessing the blasting design parameters using a box plot to remove noise and abnormal values ​​in the data samples.

[0017] A second aspect of the present invention provides a tunnel over-excavation risk assessment system based on machine learning.

[0018] A tunnel over-excavation risk assessment method based on machine learning, comprising:

[0019] A model training module is configured to: train a prediction model based on the acquired historical blasting design parameters, geological characteristics, fault crossings, tunnel geometry, and corresponding historical overbreak values; during the training process, use a quilt group optimization algorithm to obtain the best hyperparameters of the prediction model to obtain an optimized prediction model;

[0020] The prediction module is configured to obtain blasting design parameters, geological characteristics, fault crossings and tunnel geometry, and use an optimized prediction model to obtain over-excavation values ​​to evaluate the tunnel over-excavation risk level.

[0021] A third aspect of the present invention provides a computer-readable storage medium.

[0022] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps in the tunnel over-excavation risk assessment method based on machine learning as described in the first aspect above.

[0023] A fourth aspect of the present invention provides a computer device.

[0024] A computer device comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps in the tunnel over-excavation risk assessment method based on machine learning as described in the first aspect above are implemented.

[0025] Compared with the prior art, the present invention has the following beneficial effects:

[0026] The present invention establishes a random forest (RF) ensemble machine learning model optimized by a quilt swarm optimization algorithm (TSA). The RF model shows good performance on regression and classification tasks and has a high tolerance to noise and outliers. The main function of TSA is to find the optimal hyperparameters of the RF model in the solution space. The RF-TSA model is a strong solution with high accuracy to the over-digging problem. The developed model can provide more accurate over-digging values. They can be used in practice and similar projects.

[0027] The present invention uses the designed RF-TSA model to input blasting design parameters, geological characteristics and tunnel geometry parameters into the RF model to obtain the initial fault, and then uses the TSA model to optimize the initial value, that is, to use the optimization algorithm to find the best values ​​of three hyperparameters (i.e., n estimators, maximum depth and minimum sample segmentation) to minimize the generalization error of the model. In this way, the optimal hybrid model can be obtained. Thereby providing a more accurate overbreak value and improving the accuracy of tunnel overbreak risk assessment. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] The accompanying drawings in the specification, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0029] Figure 1 is a flow chart of a tunnel over-excavation risk assessment method based on machine learning shown in the present invention;

[0030] Figure 2 It is a flow chart of the capsule group optimization algorithm shown in the present invention. DETAILED DESCRIPTION

[0031] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0032] It should be noted that the following detailed descriptions are all illustrative and intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present invention belongs.

[0033] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.

[0034] It should be noted that the flowcharts and block diagrams in the accompanying drawings illustrate the possible implementation architecture, functions and operations of the methods and systems according to various embodiments of the present disclosure. It should be noted that each box in the flowchart or block diagram can represent a module, a program segment, or a part of a code, and the module, program segment, or a part of a code may include one or more executable instructions for implementing the logical functions specified in each embodiment. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in an order different from that marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, or they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the flowchart and / or block diagram, and the combination of boxes in the flowchart and / or block diagram can be implemented using a dedicated hardware-based system that performs a specified function or operation, or can be implemented using a combination of dedicated hardware and computer instructions.

[0035] Embodiment 1

[0036] The present embodiment provides a method for assessing the risk of tunnel over-excavation based on machine learning. The present embodiment uses the method applied to a server as an example for illustration. It is understandable that the method can also be applied to a terminal, and can also be applied to a system including a terminal, a server, and a server, and is implemented through the interaction between the terminal and the server. The server can be an independent physical server, or a server cluster or a distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network servers, cloud communications, middleware services, domain name services, security services CDN, and big data and artificial intelligence platforms. The terminal can be a smart phone, a tablet computer, a laptop computer, a desktop computer, a smart speaker, a smart watch, etc., but is not limited thereto. The terminal and the server can be directly or indirectly connected via wired or wireless communication, which is not limited in this application. In the present embodiment, the method includes the following steps:

[0037] Based on the historical blasting design parameters, geological characteristics, fault crossings, tunnel geometry, and corresponding historical overbreak values, the prediction model is trained. During the training process, the quilt group optimization algorithm is used to obtain the optimal hyperparameters of the prediction model and obtain the optimized prediction model.

[0038] The blasting design parameters, geological characteristics, fault crossings and tunnel geometry are obtained, and the over-excavation value is obtained using the optimized prediction model to evaluate the tunnel over-excavation risk level.

[0039] The technical solution of this embodiment is described in detail below. Figure 1 As shown:

[0040] Step 1: According to the main reasons for the fault crossing caused by blasting, namely the blasting design parameters, geological characteristics and tunnel geometry, the corresponding parameter values ​​are determined after consulting relevant information, specifically:

[0041] The input parameters selected by the model include the number of blasting holes, hole depth, total charge, blasting length, unit consumption of explosives, geological characteristics and tunnel cross-sectional area. The number of blasting holes, hole depth, total charge, blasting length and unit consumption of explosives can be obtained from the design results of each blasting operation. The tunnel cross-sectional area and fault crossing can be obtained from the measurement results of ground laser scanning technology. The geological characteristics (RMR) are calculated using the established "basic quality" (BQ) index.

[0042] Step 2: After collecting the data of the blasting operation for the first time, the collected data is preprocessed using a box plot to remove noise and outliers in the data samples. Specifically:

[0043] The box plot was used to preprocess the database and find outliers for the data samples of the entire database to prevent outliers and noise data from having a negative impact on the random forest model or reducing the model performance when collecting data from different blasting operations.

[0044] Step 3: Based on the obtained database, divide it into training set (60%), validation set (20%), and test set (20%);

[0045] Step 4: Import the model training set into the random forest model for training to obtain the original model for predicting overburst faults, specifically:

[0046] When performing RF modeling, RF will generate a certain number of decision trees, and then each decision tree will extract multiple samples from the original sample, start splitting according to the minimum leaf node mean square error until no more features are available, and then merge the prediction results of multiple decision trees to average the final result.

[0047] Step 5: Use the quilt swarm optimization algorithm to capture the best combination of the three hyperparameters in the original model, optimize the original model, and obtain the optimized model, specifically:

[0048] The RF model consists of three important hyperparameters, namely the number of decision trees in the forest, the maximum depth of the decision tree, and the minimum number of samples required to split an internal node. The quilt swarm optimization algorithm is used to capture the best combination of these three hyperparameters.

[0049] like Figure 2 As shown, the specific steps of the capsule swarm optimization algorithm include:

[0050] The first step is to avoid conflicts between search agents. When performing search work, in order to avoid conflicts between individuals, vector To calculate the new search agent position. As shown in formula (1)(2)(3):

[0051]

[0052]

[0053]

[0054] in, is gravity, is the interaction force between individuals, c1, c2 and c3 are random values ​​in the interval [0,1], 2c1 is the ocean current advection in the deep sea, P max and P min are constants, 4 and 1 respectively.

[0055] The second step is to move to the best neighbor. After eliminating the conflicts between the search agents, the search agent is guided to move to the position of the best neighbor, as shown in formula (4):

[0056]

[0057] in, is the distance from the search agent (i.e. the hook) to the target solution, is the target solution, is the position of the hook in the current iteration, t is the current iteration, r and is a random number in the interval [0,1].

[0058] The third step is to converge to the best search agent. The search agent gradually moves closer to the position of the best agent in the current iteration, as described in formula (5):

[0059]

[0060] The fourth step is group behavior. The first two best positions of the quilt are used as reference target solutions to define the group behavior of the quilt. Other search agents update their positions according to the reference target solutions. The specific group behavior is expressed as follows:

[0061]

[0062] The determination of the best hyperparameter combination in step 5 can be captured by the following steps: Initialization parameters: Initialize the population size and necessary parameters of the TSA algorithm. Fitness evaluation: Calculate the fitness value of the population and evaluate its fitness, save the best fitness value and then improve the hyperparameters; Parameter update: Update the fitness value satisfied by the hyperparameters according to the results of each iteration, so as to find the ideal hyperparameters; Suspension condition check: When the optimal fitness value no longer changes and reaches the maximum number of iterations, the optimal hyperparameters of the hybrid RF model are obtained.

[0063] Step 6: Input the validation set of step 4 into the optimization model of step 5 to verify the optimization model. Specifically:

[0064] If the verification result is poor, repeat step 5 and optimize the model. If the verification result is good, proceed to step 7.

[0065] Step 7: Input the test set of step 4 into the optimization model of step 5, test the optimization model, obtain the final hybrid random forest model, and use the model to evaluate the risk of tunnel over-excavation.

[0066] According to the tunnel construction process standards, when the output over-excavation value meets the allowable over-excavation value specified in the standards, the model determines that the tunnel over-excavation is risk-free; when the output over-excavation value exceeds the allowable over-excavation value specified in the standards, the model determines that there is risk.

[0067] Embodiment 2

[0068] This embodiment provides a tunnel over-excavation risk assessment system based on machine learning.

[0069] A tunnel over-excavation risk assessment method based on machine learning, comprising:

[0070] A model training module is configured to: train a prediction model based on the acquired historical blasting design parameters, geological characteristics, fault crossings, tunnel geometry, and corresponding historical overbreak values; during the training process, use a quilt group optimization algorithm to obtain the best hyperparameters of the prediction model to obtain an optimized prediction model;

[0071] The prediction module is configured to obtain blasting design parameters, geological characteristics, fault crossings and tunnel geometry, and use an optimized prediction model to obtain over-excavation values ​​to evaluate the tunnel over-excavation risk level.

[0072] It should be noted that the above-mentioned model training module and prediction module are the same as the examples and application scenarios implemented by the steps in Embodiment 1, but are not limited to the contents disclosed in Embodiment 1. It should be noted that the above-mentioned modules as part of the system can be executed in a computer system such as a set of computer executable instructions.

[0073] Embodiment 3

[0074] This embodiment provides a computer-readable storage medium having a computer program stored thereon. When the program is executed by a processor, the steps in the tunnel over-excavation risk assessment method based on machine learning as described in the first embodiment above are implemented.

[0075] Embodiment 4

[0076] This embodiment provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, the steps in the tunnel over-excavation risk assessment method based on machine learning as described in the above-mentioned embodiment 1 are implemented.

[0077] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of hardware embodiments, software embodiments, or embodiments combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage and optical storage, etc.) containing computer-usable program code.

[0078] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0079] These computer program instructions may also be stored in a computer readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0080] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process in the computer or other programmable device. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0081] A person skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program, and the program can be stored in a computer-readable storage medium, and when the program is executed, it can include the processes of the embodiments of the above-mentioned methods. The storage medium can be a disk, an optical disk, a read-only memory (ROM) or a random access memory (RAM), etc.

[0082] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A tunnel over-excavation risk assessment method based on machine learning, characterized in that: include: The prediction model is trained based on the acquired historical blasting design parameters, geological characteristics, fault crossings, tunnel geometry, and corresponding historical overbreak values; During the training process, the quilt swarm optimization algorithm is used to obtain the best hyperparameters of the prediction model and obtain the optimized prediction model; The blasting design parameters, geological characteristics, fault crossings and tunnel geometry are obtained, and the over-excavation value is obtained using the optimized prediction model to evaluate the tunnel over-excavation risk level.

2. The method for assessing tunnel over-excavation risk based on machine learning according to claim 1, characterized in that: The process of using the quilt swarm optimization algorithm to obtain the optimal hyperparameters of the prediction model includes: initializing the population size and necessary parameters of the quilt swarm optimization algorithm; calculating the fitness value of the population and evaluating its fitness, saving the optimal fitness value and then improving the hyperparameters; updating the fitness value satisfied by the hyperparameters according to the results of each iteration to find the ideal hyperparameters; when the optimal fitness value no longer changes and reaches the maximum number of iterations, the optimal hyperparameters of the prediction model are obtained.

3. The method for assessing tunnel over-excavation risk based on machine learning according to claim 1, characterized in that: The prediction model adopts the RF model.

4. The method for assessing tunnel over-excavation risk based on machine learning according to claim 1, characterized in that: The optimal hyperparameters of the prediction model include: the number of decision trees in the forest, the maximum depth of the decision tree, and the minimum number of samples required to split an internal node.

5. The method for assessing tunnel over-excavation risk based on machine learning according to claim 1, characterized in that: The blasting design parameters include the number of blasting holes, hole depth, total charge, blasting length and unit consumption of explosives.

6. The method for assessing tunnel over-excavation risk based on machine learning according to claim 1, characterized in that: The cross-fault and tunnel geometries were measured using terrestrial laser scanning technology.

7. The method for assessing tunnel over-excavation risk based on machine learning according to claim 1, characterized in that: After obtaining the historical blasting design parameters, the method also includes: preprocessing the blasting design parameters using a box plot to remove noise and abnormal values ​​in the data samples.

8. A tunnel over-excavation risk assessment method based on machine learning, characterized in that: include: A model training module is configured to: train a prediction model based on the acquired historical blasting design parameters, geological characteristics, fault crossings, tunnel geometry, and corresponding historical overbreak values; during the training process, use a quilt group optimization algorithm to obtain the best hyperparameters of the prediction model to obtain an optimized prediction model; The prediction module is configured to obtain blasting design parameters, geological characteristics, fault crossings and tunnel geometry, and use an optimized prediction model to obtain an overbreak value to evaluate the tunnel overbreak risk level.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps in the tunnel over-excavation risk assessment method based on machine learning are implemented as described in any one of claims 1-7.

10. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the steps in the tunnel over-excavation risk assessment method based on machine learning are implemented as described in any one of claims 1-7.

Images