Tunnel underground water flow monitoring and intelligent control method

Abstract

The invention is applicable to the technical field of tunnel monitoring, and provides a tunnel groundwater flow monitoring and intelligent control method, which comprises the following steps of: acquiring actually measured groundwater flow, surrounding rock strain data, temperature data and crack opening based on a sensor; constructing a three-dimensional dynamic permeability field model, determining a permeability change rate peak value according to the three-dimensional dynamic permeability field model, and correcting the actually measured groundwater flow based on the permeability change rate peak value to obtain a compensation correction flow; performing data cleaning, normalization and stability test on the compensation correction flow; and training and verifying the ARIMA model based on the historical compensation and correction flow data, and predicting the groundwater flow in the future time through the verified ARIMA model. The constructed three-dimensional dynamic permeability field model considers the influence of surrounding rock strain, temperature and crack propagation on the permeability, and can reflect the dynamic change of the tunnel surrounding rock permeability in real time, so that the prediction result better conforms to the actual engineering condition.

Description

Technical Field

[0001] The present invention relates to the technical field of tunnel monitoring, and in particular to a tunnel groundwater flow monitoring and intelligent control method. Background Art

[0002] During tunnel construction and operation, groundwater issues have always been a key factor affecting tunnel safety and stability. Tunnel excavation disrupts the stress balance of the existing rock mass and the groundwater seepage field, causing groundwater to seep into the tunnel. Excessive groundwater flow can trigger a series of engineering problems, such as tunnel lining leakage, softening and instability of the surrounding rock, and surface subsidence. In severe cases, it can even threaten the safety of personnel and equipment within the tunnel.

[0003] Currently, monitoring groundwater flow in tunnels primarily relies on traditional sensor technology, deploying a limited number of sensors within the tunnel to acquire real-time groundwater flow data. However, this approach has significant limitations. First, relying solely on flow monitoring data cannot fully reflect the changes in the seepage characteristics of the tunnel's surrounding rock, as these are influenced by multiple factors, such as rock strain, temperature fluctuations, and crack development. Second, most existing monitoring systems lack the ability to intelligently analyze and predict groundwater flow variations, preventing early warning of potential flow rate exceedances. This makes it difficult to implement effective control measures in the face of sudden, high-volume groundwater seepage. Furthermore, tunnel groundwater control currently primarily relies on a passive, post-processing approach, requiring only significant seepage issues to be addressed. This approach not only increases project costs and maintenance difficulties but can also cause irreversible damage to the tunnel structure. Therefore, a method for monitoring and intelligently controlling tunnel groundwater flow is needed to address these issues. Summary of the Invention

[0004] In view of the deficiencies in the prior art, the present invention aims to provide a method for monitoring and intelligently controlling groundwater flow in a tunnel, so as to solve the problems existing in the above-mentioned background technology.

[0005] The present invention is implemented as follows: a method for monitoring and intelligently controlling groundwater flow in a tunnel, the method comprising the following steps:

[0006] Based on sensor data, the measured groundwater flow, surrounding rock strain data, temperature data and fracture aperture are obtained;

[0007] Construct a three-dimensional dynamic permeability field model: K(x,y,z,t)=K0×[1+α×Δε(t) / ε0]×e -β×ΔT(t) ×(1+γ×δ(t) / δ ref), (x, y, z) represents coordinates, t represents time, K0 is the initial permeability, α is the strain coupling coefficient, Δε(t) is the real-time strain change, ε0 is the reference strain threshold, ΔT(t) is the real-time temperature deviation, β is the temperature attenuation factor, γ is the crack extension enhancement coefficient, δ(t) is the crack extension rate, δ ref is the reference crack opening;

[0008] The permeability change rate peak value is determined according to the three-dimensional dynamic permeability field model, and the measured groundwater flow is corrected based on the permeability change rate peak value to obtain the compensated corrected flow rate;

[0009] Perform data cleaning, normalization, and stationarity test on the compensated corrected flow rate to determine the orders p and q of the ARIMA model;

[0010] The ARIMA model is trained and verified based on historical compensated flow data. The verified ARIMA model is used to predict groundwater flow in the future and obtain the predicted flow value Qp(x, y, z, t+Δt).

[0011] When the predicted flow rate value is greater than the flow rate threshold, the surrounding rock deformation and crack propagation are analyzed to generate a control strategy.

[0012] As a further solution of the present invention: the sensor includes an electromagnetic flowmeter, a fiber Bragg grating strain-temperature sensor and a piezoelectric ceramic microcrack sensor. The measured groundwater flow is obtained by the electromagnetic flowmeter, the surrounding rock strain data and temperature data are obtained by the fiber Bragg grating strain-temperature sensor, and the crack aperture is obtained by the piezoelectric ceramic microcrack sensor.

[0013] As a further solution of the present invention: the initial permeability K0 is measured by indoor core tests, the temperature attenuation factor β is calibrated by thermal-fluid coupling experiments, and the crack growth rate δ(t) is obtained by time-frequency analysis of piezoelectric ceramic signals.

[0014] As a further solution of the present invention: when the measured groundwater flow is corrected based on the peak value of the permeability change rate to obtain the compensated corrected flow, a flow correction formula will be constructed: Qc(x,y,z,t)=Qm(x,y,z,t)×[1+λ×peak value of the permeability change rate], wherein Qc(x,y,z,t) is the compensated corrected flow, Qm(x,y,z,t) is the measured groundwater flow, and λ is the flow correction coefficient; the permeability value K(x,y,z,t) calculated by the three-dimensional dynamic permeability field model is differentiated in the time dimension to obtain the permeability change rate, the permeability change rate is calculated at different positions (x,y,z) of the tunnel surrounding rock, and the peak value of the permeability change rate at each position during the monitoring time period is found.

[0015] As a further solution of the present invention: the step of performing data cleaning and normalization on the compensated corrected flow rate specifically includes:

[0016] Identify outliers in the compensated flow based on statistical methods and replace them with the average value of adjacent normal data;

[0017] The compensated corrected flow rate is normalized and mapped to the [0,1] interval. The mapping formula is: Qn(x,y,z,t)=(Qc(x,y,z,t)-Qmin) / (Qmax-Qmin), where Qmin and Qmax are the minimum and maximum values ​​of the compensated corrected flow rate Qc(x,y,z,t), respectively.

[0018] As a further solution of the present invention, the steps of performing a stationarity test on the compensated corrected flow and determining the orders p and q of the ARIMA model specifically include:

[0019] The autocorrelation function, partial autocorrelation function diagram and unit root test are used to determine whether the compensation correction flow is stable, and the unsteady data are subjected to differential processing;

[0020] The orders p and q of the ARIMA model are preliminarily determined based on the autocorrelation function and partial autocorrelation function graphs of the differenced stationary data. If the autocorrelation function graph is truncated after the lag q order and the partial autocorrelation function graph is truncated after the lag p order, then the ARIMA (p, d, q) model is preliminarily determined, where d is the difference order.

[0021] The Akaike Information Criterion and the Bayesian Information Criterion are used to further determine the optimal order. By traversing different order combinations (p, q), the corresponding AIC and BIC values ​​are calculated, and the order combination that minimizes the AIC and BIC values ​​is selected as the optimal model order.

[0022] As a further solution of the present invention: when performing differential processing on unstable data, the first-order differential formula is ΔQn(x,y,z,t)=Qn(x,y,z,t)-Qn(x,y,z,t-1); if the data is still unstable after the first-order differential, the second-order differential is performed, and the second-order differential formula is Δ 2 Qn(x,y,z,t)=ΔQn(x,y,z,t)-ΔQn(x,y,z,t-1).

[0023] As a further solution of the present invention: the step of training and validating the ARIMA model based on the historical compensated and corrected traffic data specifically includes:

[0024] The historical compensation-corrected traffic data is divided into a training set and a test set in a ratio of 7:3;

[0025] Use the training set data ARIMA model to train and estimate model parameters, using maximum likelihood estimation or least squares method for parameter estimation;

[0026] Use the test set data to verify the trained model and calculate the error index between the predicted value and the actual value. If the error index meets the preset requirements, the model passes the verification; otherwise, readjust the model order or parameters, and retrain and verify.

[0027] As a further solution of the present invention: the step of analyzing surrounding rock deformation and crack propagation and generating a control strategy specifically includes:

[0028] Analyze surrounding rock deformation and crack expansion to determine whether the risk type is dominated by surrounding rock deformation or crack expansion;

[0029] When the surrounding rock deformation is dominant, the control strategy is grouting reinforcement and anchor support;

[0030] When the crack expansion is dominant, the control strategies are chemical grouting and surface plugging.

[0031] As a further solution of the present invention: the error indicators are root mean square error, mean absolute error and mean absolute percentage error.

[0032] Compared with the prior art, the present invention has the following beneficial effects:

[0033] The three-dimensional dynamic permeability field model constructed by this invention considers the impact of multiple factors on permeability, such as surrounding rock strain, temperature, and crack propagation. It can reflect the dynamic changes in tunnel surrounding rock permeability in real time, making the predictions more consistent with actual engineering conditions. It also uses a validated ARIMA model to predict future groundwater flow, providing ample time for decision-making and preparation for tunnel project operations and maintenance, effectively reducing project risks. It also analyzes surrounding rock deformation and crack propagation to generate control strategies, reducing unnecessary engineering measures and lowering project costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 The figure is a flow chart of a tunnel groundwater flow monitoring and intelligent control method.

[0035] Figure 2 The present invention is a flowchart for data cleaning and normalization in a tunnel groundwater flow monitoring and intelligent control method.

[0036] Figure 3 This is a flow chart for performing stability testing in a tunnel groundwater flow monitoring and intelligent control method.

[0037] Figure 4This is a flowchart for training an ARIMA model in a tunnel groundwater flow monitoring and intelligent control method.

[0038] Figure 5 A flow chart for generating a control strategy for a tunnel groundwater flow monitoring and intelligent control method. DETAILED DESCRIPTION

[0039] In order to make the purpose, technical solutions and advantages of the present invention clearer, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0040] The specific implementation of the present invention is described in detail below with reference to specific embodiments.

[0041] like Figure 1 As shown, an embodiment of the present invention provides a method for monitoring and intelligently controlling groundwater flow in a tunnel, the method comprising the following steps:

[0042] S100, based on sensor data, obtains measured groundwater flow, surrounding rock strain data, temperature data, and fracture aperture;

[0043] S200, construct a three-dimensional dynamic permeability field model: K(x,y,z,t)=K0×[1+α×Δε(t) / ε0]×e -β×ΔT(t) ×(1+γ×δ(t) / δ ref ), (x, y, z) represents coordinates, t represents time, K0 is the initial permeability, α is the strain coupling coefficient, Δε(t) is the real-time strain change, ε0 is the reference strain threshold, ΔT(t) is the real-time temperature deviation, β is the temperature attenuation factor, γ is the crack extension enhancement coefficient, δ(t) is the crack extension rate, δ ref is the reference crack opening;

[0044] S300, determining a permeability change rate peak value according to the three-dimensional dynamic permeability field model, and correcting the measured groundwater flow rate based on the permeability change rate peak value to obtain a compensated corrected flow rate;

[0045] S400, performing data cleaning, normalization, and stationarity test on the compensation correction flow, and determining the orders p and q of the ARIMA model;

[0046] S500: Training and validating the ARIMA model based on the historical compensated flow data, and predicting the groundwater flow in the future using the validated ARIMA model to obtain a predicted flow value Qp(x, y, z, t+Δt);

[0047] S600: When the predicted flow rate value is greater than the flow rate threshold, the surrounding rock deformation and crack extension are analyzed to generate a control strategy.

[0048] It's important to note that existing technologies primarily focus on monitoring groundwater flow itself, while ignoring the impact of key parameters such as surrounding rock strain, temperature, and fracture aperture on groundwater seepage. Changes in these parameters can directly or indirectly alter surrounding rock permeability, thereby affecting groundwater flow. The lack of comprehensive monitoring and analysis of these relevant parameters results in an incomplete and inaccurate understanding of groundwater flow variations. Traditional permeability models are typically based on static geological conditions and fail to reflect real-time permeability changes in tunnel surrounding rock during excavation and operation. In actual projects, however, surrounding rock permeability changes dynamically with factors such as stress state, temperature fluctuations, and fracture development. The lack of dynamic permeability models leads to a distorted understanding of groundwater seepage patterns, making it difficult to accurately predict groundwater flow. Most existing groundwater flow monitoring systems lack predictive capabilities, making it impossible to predict future flow trends. This makes it impossible to take preventive measures in the face of potential flow rate exceedances, forcing companies to passively address existing seepage issues, increasing project risks and costs. In terms of groundwater control, existing methods often adopt unified control measures and fail to formulate targeted control strategies based on the different causes of water seepage. For example, for water seepage problems caused by surrounding rock deformation and crack expansion, the degree of their impact on seepage is not distinguished, resulting in poor control effects and may even cause additional damage to the tunnel structure. The entire monitoring and control process mostly relies on manual operation and empirical judgment, lacking an intelligent decision support system. Manual operation is not only inefficient, but also prone to misjudgment and omission, and cannot meet the real-time and accuracy requirements of tunnel engineering. The embodiments of the present invention are intended to solve the above problems.

[0049] In the embodiment of the present invention, various sensors are first arranged, including electromagnetic flowmeters, fiber Bragg grating strain-temperature sensors, and piezoelectric ceramic microcrack sensors. The electromagnetic flowmeter is used to obtain the measured groundwater flow, the fiber Bragg grating strain-temperature sensor is used to obtain the surrounding rock strain data and temperature data, and the piezoelectric ceramic microcrack sensor is used to obtain the crack aperture. By simultaneously collecting multi-source data such as groundwater flow, surrounding rock strain, temperature, and crack aperture, the changes in the seepage characteristics of the tunnel surrounding rock can be more comprehensively and accurately reflected. The comprehensive analysis of multi-source data helps to deeply understand the intrinsic relationship between groundwater seepage and surrounding rock deformation, temperature changes, and crack development, providing a more reliable basis for subsequent flow prediction and control. Then, a three-dimensional dynamic permeability field model is constructed: K(x,y,z,t)=K0×[1+α×Δε(t) / ε0]×e -β×ΔT(t) ×(1+γ×δ(t) / δ ref), (x, y, z) represents the coordinates, t represents the time, K0 is the initial permeability, α is the strain coupling coefficient, which reflects the enhancement effect of surrounding rock deformation on permeability, Δε(t) is the real-time strain change, ε0 is the reference strain threshold, which is a constant value, ΔT(t) is the real-time temperature deviation, β is the temperature attenuation factor, γ is the crack extension enhancement coefficient, which characterizes the nonlinear contribution of microcracks to permeability, δ(t) is the crack extension rate, δ ref A reference crack aperture of 0.1 mm was used. The initial permeability, K0, was determined through indoor core testing. The temperature attenuation factor, β, was calibrated through thermal-fluid coupling experiments. The crack growth rate, δ(t), was obtained through time-frequency analysis of piezoelectric ceramic signals. This constructed a three-dimensional dynamic permeability field model that considers the influence of multiple factors on permeability, including surrounding rock strain, temperature, and crack growth. It can reflect the dynamic changes in tunnel surrounding rock permeability in real time. This model provides key technical support for accurately predicting groundwater flow, making the predictions more consistent with actual engineering conditions. Next, the three-dimensional dynamic permeability field model determines the peak permeability change rate. Based on this peak permeability change rate, the measured groundwater flow rate is corrected to obtain a compensated corrected flow rate. The compensated corrected flow rate undergoes data cleaning, normalization, and stationarity testing. The ARIMA model orders p and q are determined. The ARIMA model is then trained and validated based on historical compensated corrected flow rate data. The validated ARIMA model is used to predict future groundwater flow rates, yielding a predicted flow rate Qp(x, y, z, t+Δt). This allows for early prediction of flow rate trends over a period of time, providing sufficient time for decision-making and preparation for tunnel project operations and maintenance, effectively reducing project risks. When the predicted flow rate exceeds the flow threshold, surrounding rock deformation and crack propagation are analyzed to generate a control strategy. This personalized control approach improves control effectiveness, reduces unnecessary engineering measures, lowers project costs, and avoids additional damage to the tunnel structure. The entire monitoring and control process is intelligent, with the intelligent control system automatically completing tasks such as data collection, analysis, prediction, and control command issuance. The intelligent decision support system has greatly improved work efficiency and accuracy, reduced errors caused by human intervention, and can respond to changes in tunnel groundwater flow in real time, ensuring the safe and stable operation of tunnel projects.

[0050] In an embodiment of the present invention, when the measured groundwater flow is corrected based on the peak value of the permeability change rate to obtain the compensated corrected flow, a flow correction formula is constructed: Qc(x,y,z,t)=Qm(x,y,z,t)×[1+λ×peak value of the permeability change rate], wherein Qc(x,y,z,t) is the compensated corrected flow, Qm(x,y,z,t) is the measured groundwater flow, and λ is the flow correction coefficient, which can be obtained by fitting historical data; the permeability value K(x,y,z,t) calculated by the three-dimensional dynamic permeability field model is differentiated in the time dimension to obtain the permeability change rate, the permeability change rate is calculated at different positions (x,y,z) of the tunnel surrounding rock, and the peak value of the permeability change rate at each position during the monitoring time period is found.

[0051] like Figure 2 As shown, as a preferred embodiment of the present invention, the step of performing data cleaning and normalization on the compensation correction flow specifically includes:

[0052] S401, identifying abnormal values ​​in the compensated and corrected flow rate based on a statistical method, and replacing them with the average value of adjacent normal data;

[0053] S402: normalize the compensated flow rate and map it to the interval [0, 1].

[0054] In this embodiment of the present invention, outliers in the compensated flow rate are identified using statistical methods (such as the 3σ principle) and replaced with the average of the adjacent normal data. Furthermore, to eliminate the impact of data dimension, the compensated flow rate is normalized and mapped to the [0,1] interval using the following mapping formula: Qn(x,y,z,t) = (Qc(x,y,z,t) - Qmin) / (Qmax - Qmin), where Qmin and Qmax are the minimum and maximum values ​​of the compensated flow rate Qc(x,y,z,t), respectively.

[0055] like Figure 3 As shown, as a preferred embodiment of the present invention, the steps of performing a stationarity test on the compensated corrected flow and determining the orders p and q of the ARIMA model specifically include:

[0056] S403, using the autocorrelation function and partial autocorrelation function graph and the unit root test to determine whether the compensated flow is stable, and performing differential processing on the unstable data;

[0057] S404: Preliminarily determine the orders p and q of the ARIMA model based on the autocorrelation function and partial autocorrelation function graph of the differenced stationary data. If the autocorrelation function graph is truncated after lag q and the partial autocorrelation function graph is truncated after lag p, then preliminarily determine an ARIMA (p, d, q) model.

[0058] S405, using the Akaike Information Criterion and the Bayesian Information Criterion to further determine the optimal order, by traversing different order combinations (p, q), calculating the corresponding AIC and BIC values, and selecting the order combination that minimizes the AIC and BIC values ​​as the optimal model order.

[0059] In the embodiment of the present invention, the autocorrelation function (ACF) and partial autocorrelation function (PACF) diagrams and the unit root test are used to determine whether the compensated flow is stable. If the ACF diagram and the PACF diagram show that the data has an obvious trend or seasonality, and the ADF test statistic is greater than the critical value, it means that the data is not stable. Differentiation processing is performed on the unstable data. When performing differential processing on the unstable data, the first-order difference formula is ΔQn(x,y,z,t)=Qn(x,y,z,t)-Qn(x,y,z,t-1); if the data is still not stable after the first-order difference, the second-order difference is performed. The second-order difference formula is Δ 2 Qn(x,y,z,t)=ΔQn(x,y,z,t)-ΔQn(x,y,z,t-1). The orders p and q of the ARIMA model are then preliminarily determined based on the autocorrelation function and partial autocorrelation function graphs of the differenced stationary data. If the autocorrelation function graph is truncated after lag q and the partial autocorrelation function graph is truncated after lag p, the ARIMA(p,d,q) model is preliminarily determined, where d is the difference order. Finally, the Akaike Information Criterion and Bayesian Information Criterion are used to further determine the optimal order. By traversing different order combinations (p,q), the corresponding AIC and BIC values ​​are calculated, and the order combination that minimizes the AIC and BIC values ​​is selected as the optimal model order.

[0060] like Figure 4 As shown in FIG. 1 , as a preferred embodiment of the present invention, the steps of training and validating the ARI MA model based on the historical compensated and corrected traffic data specifically include:

[0061] S501, dividing the historical compensation-corrected traffic data into a training set and a test set;

[0062] S502, using the training set data ARIMA model to train and estimate model parameters, using maximum likelihood estimation or least squares method to perform parameter estimation;

[0063] S503, use the test set data to verify the trained model, calculate the error index between the predicted value and the actual value, if the error index meets the preset requirements, the model passes the verification; otherwise, readjust the model order or parameters, and re-train and verify.

[0064] In an embodiment of the present invention, the historical compensated and corrected traffic data is divided into a training set and a test set in a ratio of 7:3 or 8:2. The training set is used for model training, and the test set is used to evaluate the model prediction performance. The ARIMA model of the training set data is then used for training to estimate the model parameters, and the maximum likelihood estimation method or the least squares method is used for parameter estimation. Finally, the trained model is verified using the test set data, and the error index between the predicted value and the actual value is calculated. If the error index meets the preset requirements, the model passes the verification; otherwise, the model order or parameters are readjusted, and training and verification are repeated. The error indicators are root mean square error, mean absolute error, and mean absolute percentage error.

[0065] like Figure 5 As shown, as a preferred embodiment of the present invention, the step of analyzing surrounding rock deformation and crack propagation and generating a control strategy specifically includes:

[0066] S601, analyzing surrounding rock deformation and crack expansion to determine whether the risk type is surrounding rock deformation-dominated or crack expansion-dominated;

[0067] S602, when the surrounding rock deformation is dominant, the control strategy is grouting reinforcement and anchor support;

[0068] S603: When the crack expansion is dominant, the control strategies are chemical grouting and surface plugging.

[0069] In an embodiment of the present invention, the risk type will be determined to be the surrounding rock deformation-dominated type or the crack expansion-dominated type. When the surrounding rock deformation-dominated type is the control strategy, the grouting reinforcement method and the anchor support method are used. Grouting reinforcement: For areas with large surrounding rock deformation, the grouting reinforcement method is adopted. According to the geological conditions and deformation of the surrounding rock, suitable grouting materials (such as cement slurry, chemical slurry) and grouting parameters (such as grouting pressure, grouting volume) are selected. Grouting is used to fill the cracks in the surrounding rock to improve the integrity and strength of the surrounding rock and reduce the groundwater seepage channel. Anchor support: Anchors are installed on the surface of the surrounding rock to actively support the surrounding rock. The layout spacing and length of the anchors should be optimized according to the surrounding rock deformation to effectively constrain the surrounding rock deformation. When the crack expansion-dominated type is the control strategy, the chemical grouting plugging method and the surface plugging method are used. Chemical grouting plugging: For areas where cracks expand faster, chemical grouting is used to plug the cracks. Chemical grouting materials with good permeability and adhesion are selected, and the grouting materials are injected into the cracks by pressure to fill the crack space and prevent groundwater seepage. Surface sealing: Apply waterproof paint or lay waterproof membrane on the surface of the crack to form a surface sealing layer to reduce the infiltration of surface water or groundwater into the tunnel through the cracks.

[0070] The above is only a detailed description of the preferred embodiments of the present invention, which is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

[0071] It should be understood that, although the various steps in the flow chart of each embodiment of the present invention are shown in sequence according to the indication of the arrows, these steps are not necessarily performed in sequence according to the order indicated by the arrows. Unless otherwise specified herein, the execution of these steps is not strictly limited in order, and these steps can be performed in other orders. Moreover, at least a portion of the steps in each embodiment may include a plurality of sub-steps or a plurality of stages, and these sub-steps or stages are not necessarily performed at the same time, but can be performed at different times, and the execution order of these sub-steps or stages is not necessarily performed in sequence, but can be performed in turn or alternately with at least a portion of other steps or sub-steps or stages of other steps.

[0072] Those skilled in the art will appreciate that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program. The program can be stored in a non-volatile computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0073] Those skilled in the art will readily appreciate other embodiments of the present disclosure after considering the disclosure in the specification and examples. This application is intended to cover any variations, uses, or adaptations of the present disclosure that follow the general principles of the present disclosure and include common knowledge or customary techniques in the art not disclosed herein. The description and examples are to be considered merely as exemplary, and the true scope and spirit of the present disclosure are indicated by the claims.

Claims

1. A tunnel groundwater flow monitoring and intelligent control method, characterized in that: The method comprises the following steps: Based on sensor data, the measured groundwater flow, surrounding rock strain data, temperature data and fracture aperture are obtained; Construct a three-dimensional dynamic permeability field model: K(x,y,z,t)=K0×[1+α×Δε(t) / ε0]×e -β×ΔT(t) ×(1+γ×δ(t) / δ ref ), (x, y, z) represents coordinates, t represents time, K0 is the initial permeability, α is the strain coupling coefficient, Δε(t) is the real-time strain change, ε0 is the reference strain threshold, ΔT(t) is the real-time temperature deviation, β is the temperature attenuation factor, γ is the crack extension enhancement coefficient, δ(t) is the crack extension rate, δ ref is the reference crack opening; The permeability change rate peak value is determined according to the three-dimensional dynamic permeability field model, and the measured groundwater flow is corrected based on the permeability change rate peak value to obtain the compensated corrected flow rate; Perform data cleaning, normalization, and stationarity test on the compensated flow rate to determine the orders p and q of the ARIMA model; The ARIMA model is trained and verified based on historical compensated flow data. The verified ARIMA model is used to predict groundwater flow in the future and obtain the predicted flow value Qp(x, y, z, t+Δt). When the predicted flow rate value is greater than the flow rate threshold, the surrounding rock deformation and crack propagation are analyzed to generate a control strategy.

2. The tunnel groundwater flow monitoring and intelligent control method according to claim 1 is characterized in that: The sensors include an electromagnetic flowmeter, a fiber Bragg grating strain-temperature sensor, and a piezoelectric ceramic microcrack sensor. The electromagnetic flowmeter is used to obtain the measured groundwater flow, the fiber Bragg grating strain-temperature sensor is used to obtain the surrounding rock strain data and temperature data, and the piezoelectric ceramic microcrack sensor is used to obtain the crack aperture.

3. The tunnel groundwater flow monitoring and intelligent control method according to claim 2 is characterized in that: The initial permeability K0 is measured by indoor core tests, the temperature attenuation factor β is calibrated by thermal-fluid coupling experiments, and the crack growth rate δ(t) is obtained by time-frequency analysis of piezoelectric ceramic signals.

4. The tunnel groundwater flow monitoring and intelligent control method according to claim 1, characterized in that: When the measured groundwater flow is corrected based on the peak permeability change rate to obtain the compensated corrected flow, a flow correction formula is constructed: Qc(x,y,z,t)=Qm(x,y,z,t)×[1+λ×peak permeability change rate], where Qc(x,y,z,t) is the compensated corrected flow, Qm(x,y,z,t) is the measured groundwater flow, and λ is the flow correction coefficient; the permeability value K(x,y,z,t) calculated by the three-dimensional dynamic permeability field model is differentiated in the time dimension to obtain the permeability change rate, the permeability change rate is calculated at different positions (x,y,z) of the tunnel surrounding rock, and the peak permeability change rate at each position during the monitoring time period is found.

5. The tunnel groundwater flow monitoring and intelligent control method according to claim 1 is characterized in that: The step of performing data cleaning and normalization on the compensation correction flow specifically includes: Identify outliers in the compensated flow based on statistical methods and replace them with the average value of adjacent normal data; The compensated corrected flow rate is normalized and mapped to the [0,1] interval. The mapping formula is: Qn(x,y,z,t)=(Qc(x,y,z,t)-Qmin) / (Qmax-Qmin), where Qmin and Qmax are the minimum and maximum values ​​of the compensated corrected flow rate Qc(x,y,z,t), respectively.

6. The tunnel groundwater flow monitoring and intelligent control method according to claim 5, characterized in that: The steps of performing a stationary test on the compensated corrected flow and determining the orders p and q of the ARIMA model include: The autocorrelation function, partial autocorrelation function diagram and unit root test are used to determine whether the compensation correction flow is stable, and the unsteady data are subjected to differential processing; The orders p and q of the ARIMA model are preliminarily determined based on the autocorrelation function and partial autocorrelation function graphs of the differenced stationary data. If the autocorrelation function graph is truncated after the lag q order and the partial autocorrelation function graph is truncated after the lag p order, then the ARIMA (p, d, q) model is preliminarily determined, where d is the difference order. The Akaike Information Criterion and the Bayesian Information Criterion are used to further determine the optimal order. By traversing different order combinations (p, q), the corresponding AIC and BIC values ​​are calculated, and the order combination that minimizes the AIC and BIC values ​​is selected as the optimal model order.

7. The tunnel groundwater flow monitoring and intelligent control method according to claim 6, characterized in that: When performing differential processing on unstable data, the first-order differential formula is ΔQn(x,y,z,t)=Qn(x,y,z,t)-Qn(x,y,z,t-1); if the data is still unstable after the first-order differential, the second-order differential is performed, and the second-order differential formula is Δ 2 Qn(x,y,z,t)=ΔQn(x,y,z,t)-ΔQn(x,y,z,t-1).

8. The tunnel groundwater flow monitoring and intelligent control method according to claim 1, characterized in that: The steps of training and validating the ARIMA model based on the historical compensated and corrected traffic data specifically include: The historical compensation-corrected traffic data is divided into a training set and a test set in a ratio of 7:3; Use the training set data ARIMA model to train and estimate model parameters, using maximum likelihood estimation or least squares method for parameter estimation; Use the test set data to verify the trained model and calculate the error index between the predicted value and the actual value. If the error index meets the preset requirements, the model passes the verification; otherwise, readjust the model order or parameters, and retrain and verify.

9. The tunnel groundwater flow monitoring and intelligent control method according to claim 1, characterized in that: The steps of analyzing surrounding rock deformation and crack propagation and generating a control strategy specifically include: Analyze surrounding rock deformation and crack expansion to determine whether the risk type is dominated by surrounding rock deformation or crack expansion; When the surrounding rock deformation is dominant, the control strategy is grouting reinforcement and anchor support; When the crack expansion is dominant, the control strategies are chemical grouting and surface plugging.

10. The tunnel groundwater flow monitoring and intelligent control method according to claim 8, characterized in that: The error indicators are root mean square error, mean absolute error and mean absolute percentage error.

Images