Microseismic energy uncertainty analysis method and device coupled with deep-buried tunnel TBM parameters

Abstract

The application provides a microseismic energy uncertainty analysis method and device coupled with deep-buried tunnel TBM parameters, which is used for realizing modeling of a nonlinear relationship between tunnel boring machine operation parameters and microseismic energy, identification of key parameters, and uncertainty quantification of energy prediction, and also has better applicability, so that reliable basis of high quality can be provided for tunnel boring risk monitoring and decision-making work, and deep-buried tunnel engineering can better respond to microseismic events, and the engineering can be safely and reliably carried out, and has good application prospect.

Description

Technical Field

[0001] This application relates to the field of deep-buried tunnel engineering, specifically to a method and apparatus for microseismic energy uncertainty analysis of coupled deep-buried tunnel TBM parameters. Background Technology

[0002] In the construction of deep-buried tunnels, microseismic energy prediction and risk analysis are clearly of great importance. Tunnel boring machines (TBMs) have become the main mechanized construction equipment. During the tunneling process, the operating parameters of the TBM (advance speed, cutterhead torque, total thrust, penetration depth, etc.) directly reflect the interaction between the surrounding rock and the cutterhead. Simultaneously, the surrounding rock is prone to microfractures and energy release under high ground stress environments, manifesting as microseismic events.

[0003] Existing microseismic energy prediction and risk analysis mainly employ the following two types of methods:

[0004] 1) Empirical model method: Microseismic energy is calculated through empirical equations or energy-stress relationships, such as regression-based linear models or energy release theories. However, this type of method relies on artificial assumptions and is difficult to adapt to complex nonlinear geological environments.

[0005] 2) Statistical and traditional machine learning methods: linear regression, support vector machine, random forest and other methods are used to establish the mapping relationship between tunnel boring machine parameters and microseismic energy. However, such models often output point prediction results, which cannot characterize the prediction uncertainty and lack physical interpretability.

[0006] In addition, traditional models have shortcomings in the following aspects:

[0007] Poor nonlinear adaptability: Most models can only capture linear or weakly nonlinear relationships and cannot effectively express the complex energy evolution characteristics under multi-parameter coupling;

[0008] Lack of uncertainty quantification: Prediction results are usually single-value outputs and cannot provide the energy distribution range and confidence interval;

[0009] Insufficient interpretability: The model cannot reveal the mechanism by which various tunnel boring machine parameters play a role in the energy formation process.

[0010] Therefore, there is an urgent need for a new method that can comprehensively consider the nonlinear relationships of multiple parameters of tunnel boring machines, realize energy probability prediction, and quantify uncertainty, so as to improve the predictability and interpretability of microseismic risk and provide better data for deep-buried tunnel engineering. Summary of the Invention

[0011] This application provides a method and apparatus for uncertain analysis of microseismic energy coupled with TBM parameters in deep-buried tunnels. It is used to model the nonlinear relationship between the operating parameters of the tunnel boring machine and microseismic energy, identify key parameters, and quantify the uncertainty of energy prediction. It also has good applicability, which can provide high-quality and reliable basis for tunnel excavation risk monitoring and decision-making. In turn, it can enable deep-buried tunnel projects to better respond to microseismic events and promote the safe and reliable progress of the project, showing good application prospects.

[0012] Firstly, this application provides a method for microseismic energy uncertainty analysis coupled with deep-buried tunnel TBM parameters, the method comprising:

[0013] The real-time operating parameters of the tunnel boring machine and the energy of the microseismic events recorded by the corresponding microseismic monitoring system are obtained during the operation of the deep-buried tunnel project, and the two are configured as a sample pair.

[0014] A probability distribution prediction model is constructed based on sample pairs. The probability distribution prediction model is used to predict the energy distribution of corresponding microseismic events based on the real-time operating parameters input to the model.

[0015] A corresponding Bayesian posterior model is established for the probability distribution prediction model. The Bayesian posterior model is used to quantify the uncertainty of key parameters and obtain the corresponding key parameter uncertainty quantification results.

[0016] The probability distribution prediction model is used to predict the energy distribution of microseismic events, and the corresponding risk monitoring and decision analysis are determined by combining the uncertainty quantification results of key parameters, and response processing is carried out.

[0017] Secondly, this application provides a microseismic energy uncertainty analysis device coupled with deep-buried tunnel TBM parameters, the device comprising:

[0018] The acquisition unit is used to acquire the real-time operating parameters of the tunnel boring machine and the energy of the microseismic events recorded by the corresponding microseismic monitoring system during the operation of the deep-buried tunnel project, and configure the two as a sample pair;

[0019] The building unit is used to construct a probability distribution prediction model based on sample pairs. The probability distribution prediction model is used to predict the energy distribution of corresponding microseismic events based on the real-time operating parameters input to the model.

[0020] Establish a unit to build a corresponding Bayesian posterior model for the probability distribution prediction model. The Bayesian posterior model is used to quantify the uncertainty of key parameters and obtain the corresponding key parameter uncertainty quantification results.

[0021] The application unit is used to predict the energy distribution of microseismic events using a probability distribution prediction model, and to determine suitable risk monitoring and decision analysis by combining the uncertainty quantification results of relevant key parameters, and to perform response processing.

[0022] Thirdly, this application provides a processing device, including a processor and a memory, wherein a computer program is stored in the memory, and the processor executes the method provided in the first aspect of this application when it invokes the computer program in the memory.

[0023] Fourthly, this application provides a computer-readable storage medium storing a plurality of instructions adapted for loading by a processor to execute the method provided in the first aspect of this application.

[0024] From the above, it can be concluded that this application has the following beneficial effects:

[0025] For microseismic energy prediction, this application realizes the modeling of the nonlinear relationship between tunnel boring machine operating parameters and microseismic energy, identification of key parameters, and quantification of the uncertainty of energy prediction. It also has good applicability, which can provide high-quality and reliable basis for tunnel excavation risk monitoring and decision-making. In turn, it can enable deep-buried tunnel projects to better respond to microseismic events and promote the safe and reliable progress of the project, which has good application prospects. Attached Figure Description

[0026] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0027] Figure 1 This is a schematic diagram of a microseismic energy uncertainty analysis method coupled with deep-buried tunnel TBM parameters according to this application;

[0028] Figure 2 This is a schematic diagram of a microseismic energy uncertainty analysis device for coupling deep-buried tunnel TBM parameters according to this application.

[0029] Figure 3 This is a schematic diagram of one type of processing equipment used in this application. Detailed Implementation

[0030] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0031] The terms "first," "second," etc., used in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in a sequence other than that illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or device that includes a series of steps or modules is not necessarily limited to those explicitly listed, but may include other steps or modules not explicitly listed or inherent to such processes, methods, products, or devices. The naming or numbering of steps appearing in this application does not imply that the steps in the method flow must be performed in the chronological / logical order indicated by the naming or numbering. The execution order of named or numbered process steps can be changed according to the desired technical purpose, as long as the same or similar technical effect is achieved.

[0032] The module division described in this application is a logical division. In practical applications, there may be other division methods. For example, multiple modules may be combined or integrated into another system, or some features may be ignored or not executed. In addition, the coupling or direct coupling or communication connection between modules shown or discussed may be through some interfaces, and the indirect coupling or communication connection between modules may be electrical or other similar forms, none of which are limited in this application. Furthermore, the modules or sub-modules described as separate components may or may not be physically separated, may or may not be physical modules, or may be distributed in multiple circuit modules. Some or all of the modules may be selected to achieve the purpose of the solution in this application according to actual needs.

[0033] Before introducing the microseismic energy uncertainty analysis method for coupled deep-buried tunnel TBM parameters provided in this application, the background content involved in this application will be introduced first.

[0034] The microseismic energy uncertainty analysis method, apparatus, and computer-readable storage medium provided in this application for coupled deep-buried tunnel TBM parameters can be applied to processing equipment to model the nonlinear relationship between tunnel boring machine operating parameters and microseismic energy, identify key parameters, and quantify the uncertainty of energy prediction. It also has good applicability, thus providing high-quality and reliable basis for tunnel excavation risk monitoring and decision-making, thereby enabling deep-buried tunnel projects to better respond to microseismic events and promote the safe and reliable progress of the project, showing good application prospects.

[0035] The microseismic energy uncertainty analysis method for coupled deep-buried tunnel (TBM) parameters mentioned in this application can be implemented by a microseismic energy uncertainty analysis device for coupled deep-buried TBM parameters, or by different types of processing devices such as servers, physical hosts, or user equipment (UEs) that integrate such a device. The microseismic energy uncertainty analysis device for coupled deep-buried TBM parameters can be implemented in hardware or software. The UE can be a terminal device such as a smartphone, tablet, laptop, desktop computer, or personal digital assistant (PDA). The processing devices can be configured in a device cluster.

[0036] It is understandable that the proposed solution is usually based on existing data or data that has already been collected. Therefore, the processing equipment that performs the microseismic energy uncertainty analysis method coupled with the parameters of the deep-buried tunnel TBM of this application, or the processing equipment that carries the corresponding application service of the microseismic energy uncertainty analysis method coupled with the parameters of the deep-buried tunnel TBM of this application, usually only needs to meet the required data processing capabilities. The specific equipment type and equipment deployment form are quite flexible.

[0037] If the direct acquisition of existing data mentioned above is also involved, then further hardware and software adaptations are needed for the processing equipment to enable it to acquire data. For example, if it is necessary to acquire the operating parameters of a tunnel boring machine or micro-seismic energy in real time, the corresponding acquisition equipment (which is the data source and also includes the relevant control equipment itself) can be incorporated into the equipment cluster of the processing equipment. Alternatively, the processing equipment itself can be the control part of these acquisition equipment. Or, the acquisition equipment outside the processing equipment can be triggered by a third party to perform real-time data acquisition operations.

[0038] In addition, if there is a need to display the processing progress (including the processing results), the processing device itself can be configured with the required display screen (including touch screen) to display the specific content. Of course, the processing device can also display the specific content through an external display device or other devices with a display screen.

[0039] The following section introduces the microseismic energy uncertainty analysis method for coupled deep-buried tunnel (TBM) parameters provided in this application.

[0040] First, refer to Figure 1 , Figure 1 This paper illustrates a flowchart of a microseismic energy uncertainty analysis method coupled with deep-buried tunnel TBM parameters according to this application. The microseismic energy uncertainty analysis method coupled with deep-buried tunnel TBM parameters provided in this application may specifically include the following steps S101 to S104:

[0041] Step S101: Obtain the real-time operating parameters of the tunnel boring machine and the energy of the microseismic events recorded by the microseismic monitoring system during the operation of the deep-buried tunnel project, and configure the two as a sample pair;

[0042] Understandably, the initial data, or data input, of the solution in this application consists of two main aspects: the real-time operating parameters of the tunnel boring machine (TBM) and the energy of microseismic events. The former corresponds to the real-time model input, while the latter corresponds to the prediction target designed by the model.

[0043] Both of these can be collected during the execution of the proposed solution, or involve the extraction and processing of existing local or off-site data, or can be directly entered manually.

[0044] In addition, it should be noted that although there is usually only one deep-buried tunnel project, multiple situations may be involved in actual applications. In order to apply the solution of this application from the overall level, the subsequent model application process can be carried out on only one deep-buried tunnel project (usually a certain deep-buried tunnel project).

[0045] As an example, the real-time operating parameters of the tunnel boring machine collected can be denoted as: , For the i-th parameter, Given the number of parameters, the event energy recorded by the microseismic monitoring system can be denoted as: The corresponding samples can form sample pairs, which are the training samples used to train the model:

[0046] , ,

[0047] in, This is the logarithmic energy value.

[0048] Furthermore, as an exemplary embodiment, the real-time operating parameters involved in the tunnel boring machine may specifically include parameters such as cutterhead diameter, blade rotation speed, cutterhead torque, cutterhead thrust, penetration depth, propulsion speed, total thrust, uniaxial compressive strength of the bunker, rock integrity coefficient, rock abrasion coefficient, segment assembly speed, cutter wear, slag transport speed, cutter replacement time, and ventilation system status.

[0049] Understandably, the series of specific parameters exemplified here are existing parameters that may be involved in the operation and monitoring of tunnel boring machines, and therefore have not been elaborated in detail. The important thing is that they can be included in the initial scope of the real-time operation parameters involved in this application.

[0050] Furthermore, this application may also involve a preprocessing operation of judging / verifying nonlinear relationships, laying a better foundation for the subsequent construction of probability distribution prediction models.

[0051] Specifically, as an exemplary embodiment, before constructing the probability distribution prediction model based on the sample pairs in the subsequent step S102, the method of this application may further include:

[0052] For both real-time operating parameters and microseismic event energy, the corresponding Pearson correlation coefficient and mutual information (MI) are calculated. The specific quantitative formulas involved can be expressed as follows:

[0053] ,

[0054] ,

[0055] Among them, the i-th parameter in the real-time running parameters is denoted as The energy of the i-th microseismic event is denoted as . , That is, logarithmic energy value, The Pearson correlation coefficient is used. For mutual information;

[0056] If used as a measure of linear correlation Low and as a measure of nonlinear correlation When it is high, it indicates and There is a non-linear relationship between them, and they are retained as valid data;

[0057] like High and When it is low, it indicates and If a linear relationship exists, it is ignored.

[0058] It should be noted that this is not a direct statement. and By making a comparison, both can be judged by their respective threshold values ​​to determine whether the value is high or low.

[0059] When a linear relationship is determined to exist... When combining data, it can be ignored to prevent the combined data from being used in the subsequent construction of the probability distribution prediction model. The ignoring operation can involve specific operations such as adding an ignore flag, deleting data, and adjusting the data to blank / preset fixed values, and it mainly targets the real-time operating parameters of the tunnel boring machine.

[0060] This further ensures the ability of this application to express and process complex energy evolution characteristics under multi-parameter coupling in nonlinear relationships, and more effectively avoids the problem of poor nonlinear adaptability of existing methods.

[0061] Step S102: Construct a probability distribution prediction model based on the sample pairs. The probability distribution prediction model is used to predict the energy distribution of the corresponding microseismic events based on the real-time operating parameters input to the model.

[0062] Understandably, once the real-time operating parameters of the tunnel boring machine and the sample pairs configured with the energy of the microseismic events recorded by the corresponding microseismic monitoring system are obtained, the corresponding probability distribution prediction model can be constructed.

[0063] It is important to note that the model outputs an energy distribution, not a single predicted value.

[0064] As an example, the specific model output of a probability distribution prediction model, based on the complete energy probability distribution, can also involve the predicted mean and the predicted variance.

[0065] As another example, the application of the Natural Gradient Boosting (NGBoost) algorithm can be used in the process of building a probability distribution prediction model.

[0066] Furthermore, as an exemplary embodiment, a probability distribution prediction model is constructed based on sample pairs, which may specifically include:

[0067] Based on the sample pairs, the NGBoost algorithm is used to construct a probability distribution prediction model. For the probability distribution prediction model, the following assumptions can be made:

[0068] ,

[0069] in, To predict the mean, To predict variance, express Follow the mean And the variance is It follows a normal distribution.

[0070] Furthermore, this application also allows for parameter optimization of the probability distribution prediction model using the maximum likelihood method. In this case, the corresponding objective function can be specifically expressed as:

[0071] ,

[0072] in, Let be the objective function. The number of parameters for real-time operation.

[0073] Thus, in this embodiment, while constructing the probability distribution prediction model, the model training effect and model performance are further improved by optimizing the parameters of the maximum likelihood method in the loss function configuration during the model training process.

[0074] Meanwhile, to improve the interpretability of the model application to users, this application may also involve sensitivity analysis based on SHapley Additive Explanations (SHAP) in the case of constructing a probability distribution prediction model.

[0075] Correspondingly, as an exemplary embodiment, the method of this application may further include:

[0076] Based on the probability distribution prediction model, the SHAP method is used to calculate the contribution of each parameter to the prediction output, and key parameters that meet the conditions are identified. These key parameters are also used as interpretable content in the output, and their contribution values ​​are expressed as follows:

[0077] ,

[0078] in, For parameters For the predicted average marginal contribution of energy, E is the mathematical expectation. For parameters that are not part of the i-th parameter in the real-time running parameters, This is the model's predicted output.

[0079] In practice, the default approach is to identify key or sensitive parameters based on the size of the contribution, which may involve contribution size thresholds and / or references to specific contribution size rankings.

[0080] Thus, under this operation, the average marginal contribution is used. The SHAP value is used to quantitatively analyze the marginal contribution of tunnel boring machine operating parameters to energy prediction, revealing the multi-parameter coupling mechanism, realizing the physical interpretability of the model, and avoiding the problem of insufficient interpretability of existing technologies in practical applications.

[0081] Meanwhile, this application may also involve model optimization of hyperparameters; correspondingly, the method of this application may also include:

[0082] The probability distribution prediction model is optimized for hyperparameters using the Newton-Raphson-Based Optimizer (NBRO) algorithm (i.e., Newton-Raphson optimization). The hyperparameters involved may include the learning rate. Weak learners Tree depth Hyperparameters, etc.

[0083] Specifically, the iterative update formula involved in the hyperparameter optimization process can be expressed as:

[0084] ,

[0085] in, Let be the hyperparameter vector for iteration time t. Step size, It is the second derivative. It is the first derivative.

[0086] It is easy to understand that Newton-Raphson optimization can significantly accelerate the convergence speed of the model, while also effectively reducing the risk of overfitting, thereby effectively improving the model's stability and prediction accuracy.

[0087] The optimized probability distribution prediction model can also be referred to as the optimized NGBoost model.

[0088] Step S103: Establish a corresponding Bayesian posterior model for the probability distribution prediction model. The Bayesian posterior model is used to quantify the uncertainty of key parameters and obtain the corresponding key parameter uncertainty quantification results.

[0089] Understandably, this application also considers the issue of uncertainty quantification for the practical application of the probability distribution prediction model each time, so as to provide richer, more vivid and practical data references based on the energy distribution prediction results of microseismic events output by the model, combined with the uncertainty quantification results.

[0090] In response, this application can construct a corresponding Bayesian posterior model for the probability distribution prediction model built above, so that in each model application, the uncertainty quantification process can be carried out on the key parameters determined by the SHAP method above.

[0091] More specifically, the uncertainty quantification or uncertainty analysis results of key parameters can be understood as describing the reliability of the predicted results of the energy distribution of this microseismic event, the range of the reliability of the results, and the corresponding error risks of the key parameters. This can be specifically reflected by reliability probability and failure probability.

[0092] In practical engineering operations, this can help users understand the limitations and risks of model predictions more clearly than basic model predictions, and thus make more informed and appropriate decision-making responses.

[0093] Specifically, as an exemplary embodiment, a corresponding Bayesian posterior model is established for the probability distribution prediction model to quantify the uncertainty of key parameters and obtain the corresponding key parameter uncertainty quantification results, which may include:

[0094] For the probability distribution prediction model, a corresponding Bayesian posterior model is built using the probabilistic programming library PyMC5. The quantification of the uncertainty of key parameters can be specifically expressed as follows:

[0095] ,

[0096] in, To quantify the results, express Follow the mean And the variance is The normal distribution As an auxiliary factor,

[0097] The relevant prior distribution can be specifically represented as:

[0098] ,

[0099] in, Indicates input parameters Follow the mean And the variance is The normal distribution

[0100] The posterior distribution obtained using the NUTS sampling algorithm can be specifically represented as:

[0101] ,

[0102] in, Indicates that when data is observed hour, The posterior distribution is proportional to the likelihood function. With prior distribution The product of.

[0103] Based on the probability distribution prediction model, the uncertainty quantification results of key parameters are output by combining the posterior mean and 95% confidence interval.

[0104] In simpler terms, the posterior mean and the 95% confidence interval are easy to understand. They require the Bayesian posterior model to process the real-time energy distribution prediction results after the probability distribution prediction model outputs the real-time energy distribution prediction results to obtain the posterior distribution analysis results, and then further calculate them based on the posterior distribution analysis results.

[0105] It can be noted that in this setting, this application combines NGBoost with PyMC5 for the first time, realizing the extension from point prediction to distribution prediction, and is able to quantify the uncertainty range of microseismic energy.

[0106] Step S104: Use the probability distribution prediction model to predict the energy distribution of microseismic events, and combine the uncertainty quantification results of the corresponding key parameters to determine the appropriate risk monitoring and decision analysis, and then perform response processing.

[0107] Understandably, once the probability distribution prediction model and its Bayesian posterior model are constructed, they can be put into specific model application work. That is, as the deep-buried tunnel project progresses, the real-time operating parameters of the tunnel boring machine are collected as data input, the probability distribution prediction model performs corresponding microseismic event energy distribution prediction processing, and the Bayesian posterior model performs uncertainty quantification processing of key parameters based on the microseismic event energy distribution prediction results.

[0108] The predicted results of microseismic event energy distribution, the results of key parameter uncertainty quantification, and the key parameters previously determined by the SHAP method clearly have highly visualized and interpretable results, thus providing strong data support for risk monitoring and decision analysis.

[0109] Correspondingly, more suitable risk monitoring and decision analysis can be implemented to better respond to potential microseismic events during the construction of deep-buried tunnel projects, such as personnel adjustments, material scheduling, and project schedule adjustments, so that deep-buried tunnel projects can be carried out and advanced more safely and reliably.

[0110] In specific response processing, it usually involves issuing result reminders to prompt relevant personnel to review and take further concrete actions. In a few cases, the system can respond automatically through relevant terminal devices, such as issuing early warnings at the engineering site in case of emergencies.

[0111] This may involve the presentation and processing of results such as microseismic event energy distribution prediction results, key parameter uncertainty quantification results, and key parameter processing progress. Furthermore, the specific results determination and processing of risk monitoring and decision analysis are usually completed autonomously by the system under relevant autonomous processing strategies, and may also involve confirmation and adjustment by relevant personnel.

[0112] Furthermore, as can be seen from the above description of the scheme, the scheme of this application does not require the assumption of a linear relationship. In practical situations, it can be widely used for microseismic risk modeling and energy prediction analysis for different burial depths and different types of tunnel boring machines.

[0113] In conclusion, this application achieves modeling of the nonlinear relationship between tunnel boring machine operating parameters and microseismic energy, identification of key parameters, and quantification of uncertainty in energy prediction for microseismic energy prediction. It also has good applicability, which can provide high-quality and reliable basis for tunnel excavation risk monitoring and decision-making. In turn, it can enable deep-buried tunnel projects to better respond to microseismic events and promote the safe and reliable progress of the project, showing good application prospects.

[0114] The above is an introduction to the microseismic energy uncertainty analysis method for coupled deep-buried tunnel TBM parameters provided in this application. To facilitate better implementation of the microseismic energy uncertainty analysis method for coupled deep-buried tunnel TBM parameters provided in this application, this application also provides a microseismic energy uncertainty analysis device for coupled deep-buried tunnel TBM parameters from the perspective of functional modules.

[0115] See Figure 2 , Figure 2 This is a schematic diagram of a microseismic energy uncertainty analysis device coupled with deep-buried tunnel TBM parameters according to this application. In this application, the microseismic energy uncertainty analysis device 200 coupled with deep-buried tunnel TBM parameters may specifically include the following structure:

[0116] The acquisition unit 201 is used to acquire the real-time operating parameters of the corresponding tunnel boring machine and the energy of the microseismic events recorded by the corresponding microseismic monitoring system during the operation of the deep-buried tunnel project, and configure the two as a sample pair;

[0117] The building unit 202 is used to build a probability distribution prediction model based on sample pairs. The probability distribution prediction model is used to predict the energy distribution of corresponding microseismic events based on the real-time operating parameters input to the model.

[0118] Unit 203 is established to build a corresponding Bayesian posterior model for the probability distribution prediction model. The Bayesian posterior model is used to quantify the uncertainty of key parameters and obtain the corresponding key parameter uncertainty quantification results.

[0119] Application unit 204 is used to perform energy distribution prediction processing of microseismic events using a probability distribution prediction model, and to determine suitable risk monitoring and decision analysis by combining the uncertainty quantification results of relevant key parameters, and to perform response processing.

[0120] In one exemplary embodiment, real-time operating parameters include cutterhead diameter, blade rotation speed, cutterhead torque, cutterhead thrust, penetration depth, propulsion speed, total thrust, uniaxial compressive strength of the bunker, rock integrity coefficient, rock abrasion coefficient, segment assembly speed, tool wear, slag transport speed, tool change time, and ventilation system status.

[0121] In yet another exemplary embodiment, the apparatus further includes a verification unit 205, configured to:

[0122] For both real-time operating parameters and microseismic event energy, the corresponding Pearson correlation coefficient and mutual information are calculated, and the relevant quantitative formulas are expressed as follows:

[0123] ,

[0124] ,

[0125] Among them, the i-th parameter in the real-time running parameters is denoted as The energy of the i-th microseismic event is denoted as . , , The Pearson correlation coefficient is used. For mutual information;

[0126] If used as a measure of linear correlation Low and as a measure of nonlinear correlation When it is high, it indicates and There is a non-linear relationship between them, and they are retained as valid data;

[0127] like High and When it is low, it indicates and If a linear relationship exists, it is ignored.

[0128] In yet another exemplary embodiment, the construction unit 202 is specifically used for:

[0129] Based on the sample pairs, the NGBoost algorithm is used to construct a probability distribution prediction model. For the probability distribution prediction model, the following assumptions are made:

[0130] ,

[0131] in, To predict the mean, To predict variance, express Follow the mean And the variance is The normal distribution;

[0132] For the probability distribution prediction model, the parameters are optimized using the maximum likelihood method. The corresponding objective function is expressed as:

[0133] ,

[0134] in, Let be the objective function. The number of parameters for real-time operation.

[0135] In yet another exemplary embodiment, the construction unit 202 is further configured to:

[0136] Based on the probability distribution prediction model, the SHAP method is used to calculate the contribution of each parameter to the prediction output, and key parameters that meet the conditions are identified. These key parameters are also used as interpretable content in the output, and their contribution values ​​are expressed as follows:

[0137] ,

[0138] in, for For the predicted average marginal contribution of energy, E is the mathematical expectation. For parameters that are not part of the i-th parameter in the real-time running parameters, This is the model's predicted output.

[0139] In yet another exemplary embodiment, the construction unit 202 is further configured to:

[0140] The Newton-Raphson optimization algorithm was used to optimize the hyperparameters of the probability distribution prediction model. The hyperparameters involved included the learning rate. Weak learners Tree depth ;

[0141] The iterative update formula involved in the hyperparameter optimization process is expressed as follows:

[0142] ,

[0143] in, Let be the hyperparameter vector for iteration time t. Step size, It is the second derivative. It is the first derivative.

[0144] In yet another exemplary embodiment, the establishment unit 203 is specifically used for:

[0145] For the probability distribution prediction model, a corresponding Bayesian posterior model is built using the probabilistic programming library PyMC5. The quantification of the uncertainty of key parameters is as follows:

[0146] ,

[0147] in, To quantify the results, express Follow the mean And the variance is The normal distribution As an auxiliary factor,

[0148] The relevant prior distribution is represented as follows:

[0149] ,

[0150] in, Indicates input parameters Follow the mean And the variance is The normal distribution

[0151] The posterior distribution obtained using the NUTS sampling algorithm is represented as follows:

[0152] ,

[0153] in, Indicates that when data is observed hour, The posterior distribution is proportional to the likelihood function. With prior distribution The product;

[0154] Based on the probability distribution prediction model, the uncertainty quantification results of key parameters are output by combining the posterior mean and 95% confidence interval.

[0155] This application also provides a processing device from a hardware architecture perspective. As mentioned earlier, in practice, a processing device may exist as a device cluster. In this case, each device in the device cluster can also be referred to as a processing device. See [reference needed]. Figure 3 , Figure 3This diagram illustrates a structural schematic of the processing device of this application. Specifically, the processing device may include a processor 301, a memory 302, and an input / output device 303. The processor 301 executes the computer program stored in the memory 302 to implement, for example... Figure 1 The steps of the microseismic energy uncertainty analysis method coupled with deep-buried tunnel TBM parameters in the corresponding embodiment; or, when the processor 301 executes the computer program stored in the memory 302, it implements as follows: Figure 2 Corresponding to the functions of each unit in the embodiment, the memory 302 is used to store the functions executed by the processor 301 as described above. Figure 1 The computer program required for the microseismic energy uncertainty analysis method of coupling deep-buried tunnel TBM parameters in the corresponding embodiment.

[0156] For example, a computer program may be divided into one or more modules / units, one or more of which are stored in memory 302 and executed by processor 301 to complete this application. One or more modules / units may be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program in a computer device.

[0157] The processing device may include, but is not limited to, processor 301, memory 302, and input / output device 303. Those skilled in the art will understand that the illustrations are merely examples of the processing device and do not constitute a limitation on the processing device. It may include more or fewer components than illustrated, or combine certain components, or different components. For example, the processing device may also include network access devices, buses, etc., and processor 301, memory 302, input / output device 303, etc., are connected via a bus.

[0158] Processor 301 can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor. The processor is the control center of the processing device, connecting various parts of the device through various interfaces and lines.

[0159] The memory 302 can be used to store computer programs and / or modules. The processor 301 implements various functions of the computer device by running or executing the computer programs and / or modules stored in the memory 302 and by calling data stored in the memory 302. The memory 302 may mainly include a program storage area and a data storage area. The program storage area may store the operating system, application programs required for at least one function, etc.; the data storage area may store data created according to the use of the processing device, etc. In addition, the memory may include high-speed random access memory, and may also include non-volatile memory, such as hard disk, RAM, plug-in hard disk, smart media card (SMC), secure digital card (SD), flash card, at least one disk storage device, flash memory device, or other volatile solid-state storage device.

[0160] When processor 301 executes a computer program stored in memory 302, it can specifically perform the following functions:

[0161] The real-time operating parameters of the tunnel boring machine and the energy of the microseismic events recorded by the corresponding microseismic monitoring system are obtained during the operation of the deep-buried tunnel project, and the two are configured as a sample pair.

[0162] A probability distribution prediction model is constructed based on sample pairs. The probability distribution prediction model is used to predict the energy distribution of corresponding microseismic events based on the real-time operating parameters input to the model.

[0163] A corresponding Bayesian posterior model is established for the probability distribution prediction model. The Bayesian posterior model is used to quantify the uncertainty of key parameters and obtain the corresponding key parameter uncertainty quantification results.

[0164] The probability distribution prediction model is used to predict the energy distribution of microseismic events, and the corresponding risk monitoring and decision analysis are determined by combining the uncertainty quantification results of key parameters, and response processing is carried out.

[0165] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the microseismic energy uncertainty analysis device, processing equipment, and its corresponding units for coupled deep-buried tunnel TBM parameters described above can be found in, for example... Figure 1 The description of the microseismic energy uncertainty analysis method for coupling deep-buried tunnel TBM parameters in the corresponding embodiment will not be repeated here.

[0166] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be performed by instructions, or by instructions controlling related hardware. These instructions can be stored in a computer-readable storage medium and loaded and executed by a processor.

[0167] Therefore, this application provides a computer-readable storage medium storing a plurality of instructions that can be loaded by a processor to execute the present application. Figure 1 The steps of the microseismic energy uncertainty analysis method for coupling deep-buried tunnel TBM parameters in the corresponding embodiment can be found in the following example. Figure 1 The description of the microseismic energy uncertainty analysis method coupled with the parameters of the deep-buried tunnel TBM in the corresponding embodiment will not be repeated here.

[0168] The computer-readable storage medium may include: read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0169] Because of the instructions stored in the computer-readable storage medium, the present application can be executed as described above. Figure 1 The steps of the microseismic energy uncertainty analysis method coupled with the parameters of the deep-buried tunnel TBM in the corresponding embodiment can therefore achieve the results of this application. Figure 1 The beneficial effects that can be achieved by the microseismic energy uncertainty analysis method coupled with the parameters of the deep-buried tunnel TBM in the corresponding embodiment are detailed in the preceding description and will not be repeated here.

[0170] The above provides a detailed description of the microseismic energy uncertainty analysis method, apparatus, processing equipment, and computer-readable storage medium for coupled deep-buried tunnel TBM parameters provided in this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the core ideas of this application. At the same time, those skilled in the art will recognize that there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method of microseismic energy uncertainty analysis coupled with deep-buried tunnel TBM parameters, characterized in that, The method comprises: Obtaining real-time operation parameters of a corresponding tunnel boring machine in a deep-buried tunnel engineering object working process and microseismic event energy recorded by a corresponding microseismic monitoring system, and configuring the two as a sample pair; Building a probability distribution prediction model on the basis of the sample pair, the probability distribution prediction model being used for predicting corresponding microseismic event energy distribution for input real-time operation parameters of the model; Establishing a corresponding Bayesian posterior model for the probability distribution prediction model, the Bayesian posterior model being used for quantifying uncertainty of a key parameter and obtaining a corresponding key parameter uncertainty quantification result; Using the probability distribution prediction model to perform microseismic event energy distribution prediction processing, combining the corresponding key parameter uncertainty quantification result to determine adaptive risk monitoring and decision analysis, and performing response processing; The building of the probability distribution prediction model on the basis of the sample pair comprises: On the basis of the sample pair, the probability distribution prediction model is built by using an NGBoost algorithm, and the following assumptions are made for the probability distribution prediction model: ， wherein, is the predicted mean, is the predicted variance, denotes obeys a normal distribution with mean and variance . The parameters of the probability distribution prediction model are optimized by a maximum likelihood method, and a corresponding objective function is represented as: ， wherein, is the objective function, is the number of parameters of the real-time operating parameters; The establishment of the corresponding Bayesian posterior model for the probability distribution prediction model comprises: For the probability distribution prediction model, a corresponding Bayesian posterior model is established by using a probability programming library PyMC5, and quantification of uncertainty of the key parameter is represented as: ， wherein, for quantization results, denotes obeys a normal distribution with mean and variance is an auxiliary factor,​ A related prior distribution is represented as: ， wherein, represents an input parameter is subject to a normal distribution with mean and variance , A posterior distribution obtained by using an NUTS sampling algorithm is represented as: ， wherein, represents the posterior distribution of the parameter when the data is observed, is proportional to the product of the likelihood function and the prior distribution ; On the basis of the probability distribution prediction model, the key parameter uncertainty quantification result is output in combination with a posterior mean value and a 95% confidence interval.

2. The method of claim 1, wherein, The real-time operation parameters comprise a cutter head diameter, a blade rotating speed, a cutter head torque, a cutter head thrust, a penetration degree, a advancing speed, a total thrust, a single-axis compressive strength of a shield, a rock integrity coefficient, a rock abrasion coefficient, a segment assembling speed, a cutter wear degree, a residue transportation speed, a cutter changing time, and a ventilation system state.

3. The method of claim 1, wherein, Before the building of the probability distribution prediction model on the basis of the sample pair, the method further comprises: For both the real-time operation parameters and the microseismic event energy, a corresponding Pearson correlation coefficient and mutual information are calculated respectively, and a quantification formula involved is represented as: ， ， Wherein, the i-th parameter in the real-time operation parameter is denoted as , the i-th event energy of the microseismic event energy is denoted as , , is the Pearson correlation coefficient, is the mutual information; If used as a measure of linear correlation Low and as a measure of nonlinear correlation When it is high, it indicates and There is a non-linear relationship between them, and they are retained as valid data; If high and low, then it indicates a linear relationship between is ignored.

4. The method of claim 3, wherein, The method further comprises: Based on the probability distribution prediction model, a SHAP method is used to calculate a contribution value of each parameter to a prediction output, and a key parameter meeting a condition is determined, the key parameter also being used as an explainability content for output, and the contribution value is represented as: ， wherein is For the average marginal contribution to the predicted energy, E is the mathematical expectation, is a parameter in the real-time operating parameters that does not belong to the i-th parameter, is the model predicted output.

5. The method of claim 1, wherein, The method further comprises: The Newton-Raphson optimization algorithm is used for hyperparameter optimization of the probability distribution prediction model, involving hyperparameters including learning rate , weak learner , and tree depth ; An iterative update formula involved in the hyperparameter optimization process is represented as: ， wherein, is a hyperparameter vector for iteration time t, is a step size, is a second derivative, is a first derivative.

6. A device for coupling microseismic energy uncertainty analysis of deep-buried tunnel TBM parameters, characterized in that, The device comprises: An obtaining unit is configured to obtain real-time operation parameters of a corresponding tunnel boring machine in a deep-buried tunnel engineering object working process and microseismic event energy recorded by a corresponding microseismic monitoring system, and configure the two as a sample pair; A constructing unit is configured to construct a probability distribution prediction model on the basis of the sample pairs, the probability distribution prediction model being configured to predict a corresponding microseismic event energy distribution for a model input real-time operation parameter; A establishing unit is configured to establish a corresponding Bayesian posterior model for the probability distribution prediction model, the Bayesian posterior model being configured to quantify an uncertainty of a key parameter and obtain a corresponding key parameter uncertainty quantization result; An applying unit is configured to perform microseismic event energy distribution prediction processing by using the probability distribution prediction model, determine an adaptive risk monitoring and decision analysis in combination with the corresponding key parameter uncertainty quantization result, and perform response processing; The constructing unit is specifically configured to: On the basis of the sample pairs, the probability distribution prediction model is constructed by using an NGBoost algorithm, and the following assumptions are made for the probability distribution prediction model: ， wherein, is the predicted mean, is the predicted variance, denotes obeys a normal distribution with mean and variance . Parameters of the probability distribution prediction model are optimized by a maximum likelihood method, and a corresponding objective function is represented as: ， wherein, is the objective function, is the number of parameters of the real-time operating parameters; The establishing unit is specifically configured to: For the probability distribution prediction model, the Bayesian posterior model is established by using a probability programming library PyMC5, and quantization of the uncertainty of the key parameter is represented as: ， wherein, for quantization results, denotes subject to a normal distribution with mean and variance is an auxiliary factor,​ A related prior distribution is represented as: ， wherein, represents an input parameter is subject to a normal distribution with mean and variance , A posterior distribution obtained by using an NUTS sampling algorithm is represented as: ， wherein, represents the posterior distribution of the parameter when the data is observed, is proportional to the product of the likelihood function and the prior distribution . On the basis of the probability distribution prediction model, the key parameter uncertainty quantization result is output in combination with a posterior mean value and a 95% confidence interval.

7. A processing device, characterized by A processor and a memory are included, the memory stores a computer program, and the processor executes the computer program in the memory to perform the method in any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that, The computer readable storage medium stores a plurality of instructions, and the instructions are adapted to be loaded by the processor to execute the method in any one of claims 1 to 5.

Images