Device and method for measuring water inflow of extra-long water-rich tunnel based on geoacoustic characteristics

Through the water inrush measurement device based on geological acoustic characteristics, the groundwater flow of the tunnel is monitored in real time by using geological acoustic sensors and characteristic models, the problem of inaccurate water inflow prediction in traditional measurement methods is solved, and the safety and stability of special-length water-rich tunnel projects are improved. sex.

CN119087499AInactive Publication Date: 2024-12-06CHONGQING ZHONGHUAN CONSTR +1
View PDF 7 Cites 0 Cited by

Patent Information

Application Number
CN202411201729.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-29
Publication Date
2024-12-06
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

Traditional specialty water-rich tunnel water inrush measurement methods cannot provide accurate and reliable water inrush prediction data due to geological conditions and environmental limitations, resulting in difficult to ensure engineering safety and stability.

Method used

The water inrush measurement device based on geological acoustic characteristics, including a feature data acquisition unit and a data acquisition and processing unit, uses geological acoustic sensor to obtain acoustic signal data and basic data, determine the tunnel groundwater flow through geological acoustic characteristic model, and monitor and predict the water inrush in real time .

Benefits of technology

Accurate measurement of the influx of extra-length water-rich tunnels is achieved, more reliable monitoring technology is provided, ensuring the safety and stability of engineering construction, and reducing the impact of geological conditions on the measurement results.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119087499A_ABST
    Figure CN119087499A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of acoustic wave propagation and reflection characteristic application in geologic bodies, and discloses a geoacoustic characteristic-based extra-long water-rich tunnel water inflow measuring device, which comprises a characteristic data acquisition unit and a data acquisition and processing unit, and is characterized in that the characteristic data acquisition unit comprises a geoacoustic sensor; the acquisition module is used for acquiring acoustic signal data generated by tunnel underground water flowing and tunnel basic data; the data acquisition and processing unit comprises a geoacoustic feature model and is used for processing the acquired tunnel basic data and acoustic signal data; the geoacoustic feature model determines the tunnel basic data and the relationship between the acoustic signal data and the underground water flow, and finally, based on the acoustic signal data of the tunnel to be detected and the tunnel basic data obtained by the feature data obtaining unit, the predicted value of the water inflow of the tunnel is obtained through calculation. The problem that accurate data are difficult to provide in existing water inflow measurement is solved, and a more reliable water inflow monitoring technology is provided for extra-long water-rich tunnel engineering.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of application of acoustic wave propagation and reflection characteristics in geological bodies, and in particular to a device and method for measuring water inflow in a super-long water-rich tunnel based on geological acoustic characteristics. Background Art

[0002] In the construction of extra-long and water-rich tunnels, accurate measurement of water inflow is crucial to the safety and stability of the project. However, most traditional methods for measuring water inflow in extra-long and water-rich tunnels are limited by geological conditions and the environment, and cannot provide accurate and reliable water inflow prediction data, resulting in the inability of engineering personnel to timely understand the water inflow situation inside the tunnel and take appropriate measures to deal with it, thus failing to ensure the safety and stability of the tunnel project, bringing many problems to the construction of extra-long and water-rich tunnels.

[0003] At present, the industry has proposed various methods for measuring water inflow during tunnel construction. Among them, the deterministic mathematical model method is more commonly used. The numerical model in this method can analyze the inflow of groundwater into the tunnel through geological features, but the numerical model is usually a real conceptual model that needs to be based on input geological data. Due to the complex geological characteristics of the extra-long water-rich tunnel, it is difficult to obtain accurate geological data for each geological feature separately in actual application, so its prediction accuracy is far from enough. For example, the method for predicting water inflow in a tunnel in a water-rich composite stratum disclosed in the patent with application number 202110914317.6, the error between the predicted water inflow and the measured water inflow can be controlled within ±12%, which greatly improves the accuracy of the prediction results compared with the prior art, but this method mainly considers the composite stratum situation, and on the basis of assuming that the water inflow within the tunnel hole is uniformly distributed, a formula for predicting the water inflow of the tunnel through the composite stratum is given, and then the calculated water inflow data is corrected. The application environment of this formula is relatively limited. For another example, the patent with application number 201810647118.1 discloses a method for predicting tunnel water inflow. Based on the reference of different empirical formulas, a water level adjustment coefficient is introduced. The water inflow of a working tunnel is predicted, thereby predicting the tunnel water inflow under different water levels. However, if the entire tunnel is to be numerically simulated, the workload of applying this method is very huge, and the final result is an empirical prediction value. For empirical prediction data, there are too many factors that affect its accuracy, and it is difficult to guarantee the accuracy of the results. Summary of the invention

[0004] The present invention aims to provide a device and method for measuring water inflow in extra-long water-rich tunnels based on geological acoustic characteristics, so as to solve the problem that it is difficult to provide accurate data in existing water inflow measurements, and to provide a more reliable water inflow monitoring technology for extra-long water-rich tunnel projects.

[0005] To achieve the above-mentioned purpose, the present invention adopts the following technical scheme: a water inflow measurement device for a long and water-rich tunnel based on geological acoustic characteristics, including a characteristic data acquisition unit and a data acquisition and processing unit, the characteristic data acquisition unit includes a geological acoustic sensor, which is used to obtain acoustic signal data generated by the flow of groundwater in the tunnel and tunnel basic data; the data acquisition and processing unit includes a geological acoustic characteristic model, which is used to process the acquired tunnel basic data and acoustic signal data; the geological acoustic characteristic model determines the relationship between the tunnel basic data and the acoustic signal data and the groundwater flow rate, and finally, based on the acoustic signal data and tunnel basic data of the tunnel to be measured obtained by the characteristic data acquisition unit, the predicted value of the water inflow of the tunnel is calculated.

[0006] The beneficial effects of this scheme are as follows: utilizing the geological acoustic characteristics, a new type of water inflow measurement device is provided, including a characteristic data acquisition unit and a data acquisition and processing unit, the characteristic data acquisition unit captures the current acoustic signal data of the tunnel to be measured in real time, the data acquisition and processing unit collects historical tunnel water inflow data and analyzes and processes it to obtain the relationship between the geological acoustic signal and the water inflow of the tunnel, and finally based on the relationship between the geological acoustic signal and the water inflow, the predicted value of the water inflow of the tunnel to be measured can be obtained by inputting the new acoustic signal data captured by the characteristic data acquisition unit; that is, utilizing the correlation between the geological acoustic signal and the water inflow, the real-time groundwater flow of the extra-long and water-rich tunnel is monitored, and the water inflow is accurately measured, so that engineering personnel can timely understand the water inflow situation inside the tunnel, and then take corresponding measures to deal with it, so as to solve the problem that the traditional water inflow measurement method is limited by geological conditions and environment and is difficult to provide accurate data; improve the water inflow measurement accuracy in the extra-long and water-rich tunnel project, solve technical problems in measurement, and provide reliable technical support for engineering construction and management.

[0007] Preferably, it also includes a monitoring unit and an alarm unit. The monitoring unit is used to monitor the obtained water inflow prediction value; the alarm unit is used to cooperate with the monitoring unit and issue an alarm when the monitoring unit detects that the water inflow prediction value exceeds a set water inflow threshold.

[0008] Preferably, the geological acoustic characteristic model includes an acquisition module, a training module, a selection module, an update module and a prediction module, wherein the acquisition module is used to acquire historical tunnel water inrush data; the training module is used to train the geological acoustic characteristic model based on the historical tunnel water inrush data as a training set; the selection module is used to calculate the optimal hyperparameters of the trained geological acoustic characteristic model and determine the characteristic parameters associated with the prediction of water inrush volume in the historical tunnel water inrush data; the update module is used to find the outliers and clusters in each acoustic signal data and each tunnel basic data through the optimal hyperparameters to create a consistent data set, and perform data normalization processing to complete the calculation rules of the model; the prediction module is used to accept the new acoustic signal data acquired by the characteristic data acquisition unit, and calculate the predicted value of water inrush volume based on the new acoustic signal data.

[0009] Preferably, the method for measuring water inflow in a long water-rich tunnel based on geological acoustic characteristics comprises the following steps:

[0010] S210, obtaining characteristic parameters of the tunnel to be tested;

[0011] S220, calibrating the relationship between characteristic parameters and groundwater flow based on the geoacoustic characteristic model;

[0012] S230, bringing the acquired characteristic parameters of the tunnel to be tested into the geoacoustic characteristic model to calculate and obtain the predicted value of the water inflow of the tunnel.

[0013] The beneficial effects of this scheme are as follows: unlike traditional prediction methods, this scheme monitors acoustic signals in real time based on the characteristics of geological acoustics, uses devices based on geological acoustic characteristics to obtain water inflow data corresponding to different geology, and then synthesizes data sets based on the collected data sets. The data sets are used to optimize different optimal models suitable for local geological characteristics, thereby obtaining the best formula, and the relevant parameters are brought in to obtain the predicted value of water inflow. This data is more flexible, can meet different geological conditions, and reduce the risks and uncertainties associated with various phenomena in the tunnel.

[0014] Preferably, in S210, the characteristic data acquisition unit collects a variety of acoustic signal data and a variety of tunnel basic data;

[0015] The data acquisition and processing unit processes and extracts features from the collected acoustic signal data and basic data of each tunnel to obtain a variety of geological acoustic features;

[0016] The geoacoustic features are subjected to dimensionality reduction operations to finally obtain characteristic parameters, and the dimensionality reduction operation specifically uses a stepwise method to minimize the data dimension.

[0017] Preferably, when performing data normalization processing and completing the model calculation rules, the data normalization method is to process all data in the range of 0 to 1 using the Min-Max normalization method.

[0018] Preferably, when establishing the geoacoustic characteristic model, the following steps are included:

[0019] S310, training the geoacoustic characteristic model based on the historical tunnel water inrush data collected by the data acquisition and processing unit to obtain an optimal model;

[0020] S320, determining an equation for water inflow prediction based on the optimal model;

[0021] S330, evaluating the best model to ensure the accuracy of the best water inflow prediction equation.

[0022] Preferably, the geoacoustic characteristic model uses a GEP algorithm to calculate the water inflow.

[0023] Preferably, if the tunnel is a drill-and-blast tunnel, the characteristic parameters include tunnel depth, groundwater level, rock quality index and water production rate.

[0024] Preferably, the water inflow prediction equation is:

[0025]

[0026] Wherein: H represents the tunnel depth, h1 represents the groundwater level, RQD represents the rock quality index, W represents the water production rate, Gene1, Gene2, Gene3, Gene4, Gene5, Gene6, Gene7, and Gene8 represent 8 models with different accuracies. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 is a structural block diagram of a measuring device in an embodiment of the present invention;

[0028] Figure 2 is a flow chart of a measurement method in an embodiment of the present invention;

[0029] Figure 3 Schematic diagram of the process of training a geoacoustic feature model in an embodiment of the present invention;

[0030] Figure 4 It is a scatter plot of the measured and predicted water inflow of the training set and the test set in the embodiment of the present invention. DETAILED DESCRIPTION

[0031] The following is further described in detail through specific implementation methods:

[0032] Example

[0033] The embodiment is basically as shown in the attached Figure 1-4 As shown, Figure 1 The device for measuring water inflow in a long and water-rich tunnel based on geological acoustic characteristics shown in the figure comprises a characteristic data acquisition unit 110 and a data acquisition and processing unit 120. The characteristic data acquisition unit 110 captures the current acoustic signal data of the tunnel to be measured in real time. The data acquisition and processing unit 120 collects historical tunnel water inflow data and performs analysis and processing to obtain the relationship between the geological acoustic signal and the water inflow of the tunnel. Finally, based on the relationship between the geological acoustic signal and the water inflow, the predicted value of the water inflow of the tunnel to be measured can be obtained by inputting the new acoustic signal data captured by the characteristic data acquisition unit 110 into the input relationship; that is, by utilizing the correlation between the geological acoustic signal and the water inflow, the real-time groundwater flow condition of the long and water-rich tunnel is monitored, and the water inflow is accurately measured, so that engineering personnel can timely understand the water inflow condition inside the tunnel, and then take corresponding measures to deal with it, thereby solving the problem that the traditional water inflow measurement method is limited by geological conditions and environment and is difficult to provide accurate data, providing a more reliable water inflow monitoring technology for the long and water-rich tunnel project, and ensuring that the tunnel project construction can be carried out safely and stably.

[0034] First, a characteristic data acquisition unit 110 is provided in the measuring device, which is used to obtain the acoustic signal data generated by the flow of groundwater in the tunnel and the basic data of the tunnel. The characteristic data acquisition unit 110 includes a geoacoustic sensor, which has the characteristics of high sensitivity and high frequency response. After being installed on the tunnel wall, the acoustic wave signal generated by the flow of groundwater can be captured in real time, and the geoacoustic characteristics including the groundwater level and the groundwater production rate can be obtained through the groundwater flow acoustic wave signal. At the same time, the characteristic data acquisition unit 110 also includes related instruments that can collect data such as tunnel depth, tunnel length, rock quality indicators in the tunnel, etc., to further obtain other basic tunnel data that may affect the tunnel water inflow in addition to the acoustic signal data, so as to improve the accuracy of the device in predicting the tunnel water inflow.

[0035] The data acquisition and processing unit 120 acquires and stores historical water inrush data of multiple tunnels, i.e., historical tunnel water inrush data. The data acquisition and processing unit 120 includes a geoacoustic characteristic model established based on the historical tunnel water inrush data. The geoacoustic characteristic model processes the acquired tunnel basic data and acoustic signal data to obtain characteristic parameters, and determines the relationship between the tunnel basic data and the acoustic signal data and the groundwater flow, and finally calculates the predicted value of the water inrush volume of the tunnel based on the acoustic signal data of the tunnel to be tested acquired by the characteristic data acquisition unit 110, thereby realizing the measurement of the water inrush volume.

[0036] At the same time, if Figure 1As shown, the measuring device also includes a monitoring unit and an alarm unit 140. The monitoring unit is used to monitor the predicted water inflow value output by the data acquisition and processing unit 120 in real time, and to evaluate in real time whether the predicted water consumption value output by the data acquisition and processing unit 120 exceeds the water inflow threshold by setting the safe water inflow threshold limit. The water inflow threshold can be set according to the actual situation of different tunnels. The alarm unit 140 is used to cooperate with the monitoring unit and issue an alarm when the monitoring unit detects that the predicted water inflow value exceeds the safe water inflow threshold limit, so as to remind the engineering personnel to respond in time and take corresponding safety measures.

[0037] The prediction of water inflow usually needs to be targeted according to the specific engineering geological conditions, topographical features and hydrological boundary conditions of the project. Due to the complex movement of groundwater in mountain tunnels, the existing prediction methods are far from accurate enough in terms of actual application, especially for long and water-rich tunnels. There are long-term technical difficulties in predicting water inflow. To this end, this scheme innovatively proposes a measuring device that can monitor geoacoustic signals in real time, so as to study the relationship between geoacoustic characteristic signals and tunnel water inflow, so as to calculate the accurate tunnel water inflow prediction value in real time based on the acoustic signal data obtained in real time from the current tunnel to be tested.

[0038] Applied to the above-mentioned measuring device, this embodiment also provides a method for measuring water inflow in a long water-rich tunnel based on geological acoustic characteristics, such as Figure 2 As shown, the specific steps include:

[0039] S210: Acquire characteristic parameters of the tunnel to be tested based on the measuring device.

[0040] After collecting various acoustic signal data generated by the groundwater flow in the tunnel to be tested and multiple tunnel basic data obtained by various measuring instruments through the geoacoustic sensor, the data acquisition and processing unit 120 processes and extracts features of the collected acoustic signal data and each tunnel basic data to obtain a variety of geoacoustic features; then, these geoacoustic features are reduced in dimension to finally obtain characteristic parameters.

[0041] In this step, the dimensionality reduction operation is specifically to minimize the data dimension using a stepwise method. The stepwise method simplifies the high-dimensional data by performing dimensionality reduction processing on multiple data to facilitate the use of supervised learning in the next step, that is, when processing a large number of feature data sets, by reducing irrelevant features, by observing which variables are selected to infer which variables (i.e., the feature parameters in this embodiment) are most important to the model, have the closest relationship with the target variable, and contribute the most to the model's predictive ability, thereby achieving the purpose of optimizing the model and improving the accuracy and efficiency of the prediction.

[0042] The tunnel application area of ​​this embodiment is located in the east wing trough or near the core. The rock type is mainly mud limestone and breccia with a small amount of shale. The tunnel is obtained by drilling and blasting. It is located in the deep circulation zone, near the contact zone between the first section of the Badong Formation and the second section of the Badong Formation and the fourth section of the Jialingjiang Formation. It is very easy to produce water gushing and water inrush. Finally, after minimizing the data dimension processing, there are four most important characteristic parameters for water gushing, namely: tunnel depth H, groundwater level h1, rock quality index RQD and water production rate W.

[0043] S220, based on the geoacoustic characteristic model, calibrate the relationship between the characteristic parameters and the groundwater flow.

[0044] In this embodiment, the geoacoustic feature model includes an acquisition module, a training module, a selection module, an update module and a prediction module, wherein:

[0045] The acquisition module is used to obtain historical tunnel water inrush data. The historical tunnel water inrush data is the water inrush volume data of different tunnels calculated by the same method as this scheme. It is obtained by collecting and counting the water inrush volume of previously constructed tunnels, and the data is input into the acquisition module and stored.

[0046] The training module is used to train the geoacoustic feature model based on historical tunnel water inrush data as a training set.

[0047] A selection module is used to calculate the optimal hyperparameters of the trained geoacoustic feature model and determine the characteristic parameters associated with water inrush prediction in historical tunnel water inrush data.

[0048] The optimal hyperparameters are obtained by minimizing the dimension of the data by focusing on the most important attributes in a stepwise method, and using three important indicators, including Akaike Information Criterion AIC, Akaike Information Criterion's corrected version AICc, and Bayesian Information Criterion BIC, to evaluate the model prediction errors of these values. The lower the metric value of the data, the higher the accuracy of the parameters incorporated into the model. The first four are selected as the optimal hyperparameters, namely the four characteristic parameters determined in this embodiment as recorded above: tunnel depth H, groundwater level h1, rock quality index RQD, and water production rate W. When determining the optimal hyperparameters, the difference threshold between the output data of the geoacoustic characteristic model and the real data is set to be less than 5%, which is the best.

[0049] The update module is used to find the outliers and clusters in each acoustic signal data and each tunnel basic data through the optimal hyperparameters to create a consistent data set, and perform data normalization to complete the calculation rules of the model.

[0050] Based on the optimal hyperparameters, outliers in each acoustic signal data and each tunnel basic data are found. Specifically, this embodiment finds that some parameters of H and RQD are abnormal, finds outliers in the data and clusters to create a consistent data set, and normalizes the data. The specific data normalization method is to use the Min-Max normalization method to process all data (including training and test data) in the range of 0 to 1. The normalized equation is as follows:

[0051]

[0052] Where: x represents the original data, x (0,1) Represents the normalized version of the original data, Min represents the minimum value of the entire data set, and Max represents the maximum value of the entire data set.

[0053] The normalization method maps the data to the range of [0,1]. Normalization converts the data features to the same scale to unify the data scale and solve the problem that different feature data have different sources and measurement units and large scale differences. Different scales will cause the gradient direction at most positions to not be the optimal search direction. Normalization can improve this, thereby further reducing the difficulty of data training, reducing manual intervention, and improving the accuracy and convergence speed of the model.

[0054] The prediction module is used to receive the new acoustic signal data acquired by the characteristic data acquisition unit 110, and calculate the water inflow prediction value based on the new acoustic signal data.

[0055] In this embodiment, the geological acoustic characteristic model is specifically based on the GEP algorithm to calculate and predict the water inflow. The GEP model is superior to traditional algorithms in speed and effectiveness, and has the characteristics of no need for derivation and strong model interpretability, and is suitable for solving various complex problems.

[0056] S230, bringing the acquired characteristic parameters of the tunnel to be tested into the geoacoustic characteristic model to calculate and obtain the predicted value of the water inflow of the tunnel.

[0057] By utilizing the established geological acoustic characteristic model and combining it with the new acoustic signal characteristics of the tunnel to be tested collected in real time, the GEP algorithm can be used to calculate and predict the water inflow.

[0058] like Figure 3 As shown in the figure, when establishing the geoacoustic characteristic model, the model needs to be trained through historical tunnel water inrush data, which specifically includes the following steps:

[0059] S310, training the geoacoustic characteristic model based on the historical tunnel water inrush data collected by the data collection and processing unit 120 to obtain an optimal model.

[0060] After training and evaluating to get the best model, the first step in preparing to use GEP is to select the mother function. The selection of the mother function is specifically measured by the mean square error to measure the fitness (f i ), the equation for the mean square error is as follows:

[0061]

[0062] Where: n represents the number of samples, yi represents the true value, Represents the predicted value.

[0063] The mean square error (MSE) takes the sample mean as an unbiased estimate of the population mean, treats positive and negative errors equally, and handles larger errors disproportionately, reflecting the degree of difference between the estimator and the estimated quantity. This scheme uses the mean square error as a loss function and updates the model parameters by calculating the mean square error to minimize the difference between the predicted value and the true value.

[0064] Next, the second major step involves selecting a set of terminals (T) and a set of functions (F) to create the equations for the estimation:

[0065] Functionset={+,-,\ast, / ,sin,cos}

[0066] In this embodiment, a total of 600 historical tunnel water inrush data were obtained, and the 600 data sets were divided into a training part and a test part. That is, 480 data sets of the entire database were used to train the geoacoustic feature model (480 training sets), and the remaining 120 data sets were used to evaluate the performance of the geoacoustic feature model (120 test sets).

[0067] In order to achieve the best combination of GEP parameters, multiple models with different parameter values ​​were executed based on a trial and error method. A total of eight models were selected here to provide geoacoustic signature models with higher accuracy. In order to obtain the best results, multiple runs of GEP modeling were performed. At each stage of model implementation, different values ​​of the parameters "population (n_pop)" and "number of generations (n-gen)" were considered. Finally, the best geoacoustic signature model was finally determined by inspecting the results obtained and evaluating its accuracy through multiple statistical criteria. The results of the GEP model are very sensitive to two parameters (n_pop and n_gen) and must be adjusted for the specific data set. Therefore, it is necessary to repeat the GEP model with different values ​​of these parameters.

[0068] S320, determining an equation for predicting water inflow based on the optimal model.

[0069]

[0070]

[0071] Where: H represents tunnel depth, h1 represents groundwater level, RQD represents rock quality index, W represents water production rate, Gene1, Gene2, Gene3, Gene4, Gene5, Gene6, Gene7, Gene8 represent 8 models with different precisions. It should also be noted that the calculation equation for water inflow prediction is applicable to the prediction and calculation of tunnel water inflow obtained by all drilling and blasting methods. For other types of tunnels, there may be different optimal hyperparameters in characteristic parameters, which need to be determined according to the characteristic parameters obtained after dimensionality reduction processing of the data according to the specific steps recorded above.

[0072] S330, evaluating the best model to ensure the accuracy of the best water inflow prediction equation.

[0073] This step is result analysis and model evaluation. The generated equations are tested using the input parameters in the training and test sets, and their accuracy is measured using a variety of statistical measures, i.e., the performance of the prediction model is evaluated using a variety of statistical measures, such as standard deviation and correlation coefficient. Since the above calculation equations contain all the input parameters, the average equation is more reliable and effective in estimating the tunnel water inflow.

[0074] This embodiment further proposes a 20 -index equation is used to evaluate the reliability of the geoacoustic characteristic model. The specific equation is as follows:

[0075]

[0076] Where: M represents the total number of samples in the data set, m 20 Represents the number of samples where the experimental or predicted rate value is between 0.80 and 1.20.

[0077] It should be noted that for a perfect prediction model, a 20 -index should be equal to unity. Proposed 20 -index has important advantages in physical engineering, and the proportion of samples that meet the expected value with a margin of error of 20% or less is reported using a 20 - Index, the recommended model (i.e. Eq. ) can accurately predict the water inflow of drill-and-blast tunnels. 20 -index result is as follows Figure 4 The scatter plot is shown.

[0078] Figure 4In the figure, the left side is the point distribution diagram of the water inflow measured by the training set, and the right side is the point distribution diagram of the water inflow measured by the test set. It can be seen that the points distributed in the left and right figures are relatively concentrated and overlap, and only a few samples (i.e., the scattered points on the right side of the left figure) have a deviation greater than 20%. Therefore, the equation formed by this model has potential ability in estimating tunnel water inflow and can be successfully predicted.

[0079] In this scheme, a water inflow prediction device is established by installing a geoacoustic sensor on the tunnel wall to accurately capture the acoustic signal generated by groundwater flow, and a data acquisition and processing unit 120 is designed to receive and process the data collected by the geoacoustic sensor, thereby obtaining a module, a training module, a selection module, an update module, and a training module based on the data set of historical tunnel water inflow data to obtain the best model, and finally obtain a geoacoustic characteristic model based on the GEP algorithm. The geoacoustic characteristic model identifies the relationship between each input acoustic signal data and the basic tunnel data and the water inflow, finds the best match between the geoacoustic characteristic data and each model, and creates a complete calculation equation by averaging the equations provided by the GEP algorithm. The four characteristic parameters in the data set are used as input values ​​to accurately estimate the water inflow of the tunnel, so as to predict the future water inflow of the tunnel in real time before the tunnel is built, saving time and money, and paving the way for the progress and development of water inflow prediction in drill and blast tunnels.

[0080] Different from the traditional prediction method, this scheme monitors the acoustic signal in real time based on the characteristics of geoacoustics, obtains the water inflow data corresponding to different geology based on the device of geoacoustic characteristics, and then synthesizes the data set according to the collected data set. The data set is used to optimize the different best models suitable for the local geological characteristics, so as to obtain the best formula, and the relevant parameters are brought in to obtain the predicted value of water inflow. This data is more flexible, can meet different geological conditions, and reduce the risks and uncertainties associated with various phenomena in the tunnel.

[0081] The above is only an embodiment of the present invention, and the common knowledge such as the known specific technical solutions and / or characteristics in the solution is not described in detail here. It should be pointed out that for those skilled in the art, without departing from the technical solution of the present invention, several modifications and improvements can be made, which should also be regarded as the protection scope of the present invention, and these will not affect the effect of the implementation of the present invention and the practicality of the patent. The scope of protection required by this application shall be based on the content of its claims, and the specific implementation methods and other records in the specification can be used to interpret the content of the claims.

Claims

1. A water inflow measurement device for a long and rich water tunnel based on geological acoustic characteristics, characterized by: It includes a feature data acquisition unit and a data acquisition processing unit. The characteristic data acquisition unit includes a geoacoustic sensor for acquiring acoustic signal data generated by underground water flow in the tunnel and tunnel foundation data; The data acquisition and processing unit includes a geological acoustic characteristic model, which is used to process the acquired tunnel basic data and acoustic signal data; The acoustic characteristic model determines the relationship between the tunnel basic data and the acoustic signal data and the groundwater flow rate. Finally, based on the acoustic signal data and tunnel basic data of the tunnel to be tested obtained by the characteristic data acquisition unit, the predicted value of the water inflow of the tunnel is calculated.

2. The device for measuring water inflow in a long water-rich tunnel based on geological acoustic characteristics according to claim 1 is characterized in that: It also includes a monitoring unit and an alarm unit. A monitoring unit, used for monitoring the obtained water inflow prediction value; The alarm unit is used to cooperate with the monitoring unit and issue an alarm when the monitoring unit detects that the predicted value of the water inflow exceeds a set water inflow threshold.

3. The device for measuring water inflow in a long water-rich tunnel based on geological acoustic characteristics according to claim 2 is characterized in that: The geoacoustic feature model includes acquisition module, training module, selection module, update module and prediction module. An acquisition module is used to obtain historical tunnel water inrush data; A training module is used to train the geoacoustic feature model based on historical tunnel water inrush data as a training set; A selection module is used to calculate the optimal hyperparameters of the trained geoacoustic feature model and determine the characteristic parameters associated with water inflow prediction in the historical tunnel water inflow data; The update module is used to find the outliers and clusters in each acoustic signal data and each tunnel basic data through the optimal hyperparameters to create a consistent data set, and perform data normalization to complete the calculation rules of the model; The prediction module is used to receive the new acoustic signal data acquired by the characteristic data acquisition unit, and calculate the water inflow prediction value based on the new acoustic signal data.

4. A method for measuring water inflow in a long water-rich tunnel based on geological acoustic characteristics, characterized in that: The following steps are included: S210, obtaining characteristic parameters of the tunnel to be tested; S220, calibrating the relationship between characteristic parameters and groundwater flow based on the geoacoustic characteristic model; S230, bringing the acquired characteristic parameters of the tunnel to be tested into the geoacoustic characteristic model to calculate and obtain the predicted value of the water inflow of the tunnel.

5. The method for measuring water inflow in a long water-rich tunnel based on geological acoustic characteristics according to claim 4 is characterized in that: In S210, the characteristic data acquisition unit collects a variety of acoustic signal data and a variety of tunnel basic data; The data acquisition and processing unit processes and extracts features from the collected acoustic signal data and basic data of each tunnel to obtain a variety of geological acoustic features; The geoacoustic features are subjected to dimensionality reduction operations to finally obtain characteristic parameters, and the dimensionality reduction operation specifically uses a stepwise method to minimize the data dimension.

6. The method for measuring water inflow in a long water-rich tunnel based on geological acoustic characteristics according to claim 5 is characterized in that: When performing data normalization and completing model calculation rules, the data normalization method is to process all data in the range of 0 to 1 using the Min-Max normalization method.

7. The method for measuring water inflow in a long water-rich tunnel based on geological acoustic characteristics according to claim 6 is characterized in that: The following steps are included when establishing a geoacoustic feature model: S310, training the geoacoustic characteristic model based on the historical tunnel water inrush data collected by the data acquisition and processing unit to obtain an optimal model; S320, determining an equation for water inflow prediction based on the optimal model; S330, evaluating the best model to ensure the accuracy of the best water inflow prediction equation.

8. The method for measuring water inflow in a long water-rich tunnel based on geological acoustic characteristics according to claim 7 is characterized in that: The geoacoustic characteristic model uses the GEP algorithm to calculate the water inflow.

9. The method for measuring water inflow in a long water-rich tunnel based on geological acoustic characteristics according to claim 8 is characterized in that: If the tunnel is a drill-and-blast tunnel, the characteristic parameters include tunnel depth, groundwater level, rock quality index and water production rate.

10. The method for measuring water inflow in a long water-rich tunnel based on geological acoustic characteristics according to claim 9 is characterized in that: The water inflow prediction equation is: Wherein: H represents the tunnel depth, h1 represents the groundwater level, RQD represents the rock quality index, W represents the water production rate, Gene1, Gene2, Gene3, Gene4, Gene5, Gene6, Gene7, and Gene8 represent 8 models with different accuracies.

Citation Information

Patent Citations

  • Seismic detection system of detecting underground water level

    CN103760591A

  • Real-time and online in-situ underground water monitoring system

    CN110208847A

  • Mine water inflow prediction method and device

    CN115983488A

  • Tunnel water inflow monitoring method in metamorphic rock area

    CN117268499A

  • Tunnel water inflow intelligent monitoring system and early warning method

    CN117345342A