Tunnel hydraulic transient coupling support control system and method based on narrowband Internet of Things

Abstract

The invention discloses a tunnel hydraulic transient coupling support control system and a tunnel hydraulic transient coupling support control method based on a narrowband internet of things, and relates to the technical field of underground engineering safety monitoring. Synchronous monitoring of a single anchor rod node in multiple physical quantity dimensions is realized. Particularly, polar angle parameters of a three-dimensional space are obtained through the micro inertial measurement unit and the laser ranging module, so that each disturbance signal has an accurate space positioning label, and the limitation that traditional monitoring is only in a time domain or a scalar signal layer is broken through; and a high-resolution space reference dimension is provided for subsequently constructing a disturbance direction and a propagation path. A normalized cross-correlation function is calculated based on a water stress independent signal set, and the response lag relation between water pressure disturbance and anchor rod strain is effectively described. In the process, the actual delay of disturbance propagation among the anchor points can be dynamically identified by introducing an integral window and average offset correction.

Description

Technical Field

[0001] This invention relates to the field of underground engineering safety monitoring technology, specifically to a tunnel hydraulic transient coupling support control system and method based on narrowband Internet of Things. Background Technology

[0002] During tunnel excavation and operation, there is a complex interaction between the pore hydrodynamic disturbance within the rock mass and the anchor support system. Especially under conditions of sudden water inrush, unloading disturbance, or abrupt changes in geological stress, hydraulic transient waves can induce stress redistribution in the surrounding rock, thereby affecting the stress state and structural stability of the anchors. Therefore, identifying the anchor response behavior and establishing a self-regulation mechanism driven by hydraulic disturbance has become one of the key issues in the intelligent evolution of tunnel support systems.

[0003] Currently, most commonly used support structure monitoring methods rely on independent distributed sensors to collect anchor strain data or single-point pore pressure sensors to monitor surrounding rock water pressure. Their design philosophy emphasizes "independent anomaly identification of a single physical quantity," but fails to construct a dynamic coupling model between hydraulic disturbance and structural response. These methods suffer from drawbacks such as fragmented monitoring dimensions, weak disturbance source tracing capabilities, and large response delays. They struggle to accurately characterize the propagation path and response link between "water-rock-anchor" in the unloading zone, resulting in a system that can only passively identify damage and cannot achieve active control.

[0004] Pore ​​water pressure in surrounding rock is prone to high-frequency transient fluctuations under stress disturbance. If its propagation path and scope of influence are not identified in time, it can easily lead to abnormal phenomena such as prestress imbalance of unloaded anchor bolts and local support failure. If it continues to develop, it will lead to response mismatch of anchor bolt groups, decreased stability of the support system, and in severe cases, may induce engineering risk events such as surrounding rock collapse, water inrush, or shear failure of support components. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a tunnel hydraulic transient coupling support control system and method based on narrowband Internet of Things, which solves the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solution: a tunnel hydraulic transient coupling support control system based on narrowband Internet of Things, including a data acquisition and processing module, a signal decoupling and feature extraction module, a transient delay calculation and stress propagation direction matrix construction module, a collaborative response determination module, and a tension control execution module; The data acquisition and processing module collects and preprocesses the tunnel hydraulic data through sensors installed between each anchor bolt and the surrounding rock to obtain the coupled dataset UW. The signal decoupling and feature extraction module decouples the coupled dataset UW, extracts the propagation characteristics of hydraulic transient waves, calculates the correlation between water pressure disturbance and strain response, and obtains the normalized cross-correlation function Cwp. The transient delay calculation and stress propagation direction matrix construction module is based on the normalized cross-correlation function Cwp. It performs spatial interpolation on the measurement points of the anchor bolt, calculates the spatial gradient, forms the stress propagation direction matrix D, and obtains the direction vector intensity Mdir. The collaborative response determination module uses the stress propagation direction matrix D and the direction vector intensity Mdir as a basis to perform a directional coupling determination function on the edge nodes and calculate the collaborative response index Rc of the nodes. The tension control execution module combines the stress propagation direction matrix D and the cooperative response index Rc to obtain the set of anchor bolts Ωadj that need to be corrected, and calculates the compensation tension force ΔF.

[0007] The data acquisition and processing module includes a multi-source signal collaborative acquisition unit and a signal denoising and normalization fusion unit; The multi-source signal collaborative acquisition unit collects data on the axial micro-strain Ax of the anchor bolt, pore water pressure Pw, transient water pressure disturbance value Ps, spatial polar angle the of the sensing point, and pore water density ps of the surrounding rock through sensors installed between each anchor bolt and the surrounding rock. Among them, the axial micro-strain Ax of the anchor bolt is acquired by a fiber optic strain sensor array; The axial micro-strain Ax of the anchor bolt is obtained by the following formula: ; In the formula, ΔλB represents wavelength drift, λB represents the initial Bragg wavelength, and KFBG represents the calibration coefficient; The pore water pressure Pw is acquired by a pore water pressure sensor. The method for acquiring pore water pressure Pw is as follows: First, the pore water pressure sensor is installed at the tail of the anchor bolt or inside the grouting borehole. Then, the system periodically excites the vibrating string, and the frequency is measured by self-excitation and attenuation. The frequency is then converted into a water pressure value to obtain the pore water pressure Pw. The spatial polar angle the of the sensing point is acquired through a miniature inertial measurement unit and a laser ranging module; The original dataset YW was obtained by fitting the acquired axial micro-strain Ax of the anchor bolt, pore water pressure Pw, and spatial polar angle the of the sensing point. Preferably, the signal denoising and normalization fusion unit preprocesses the original dataset YW to obtain the coupled dataset UW; Preprocessing includes data time alignment and synchronization, noise reduction, and normalization. Data time alignment and synchronization are achieved by using an adaptive time remapping algorithm based on a local self-similar kernel function to align the data time of the original dataset YW; this improves the synchronization accuracy of multi-channel signals and can dynamically adjust the time axis scaling of multi-channel signals, thereby correcting minute sampling drift and ensuring that strain and water pressure signals strictly correspond within a millisecond time window; Denoising identifies anomalous noise segments by comparing residual spectral density. First, the spectral residual function of the signal is calculated: The spectral residual function is obtained as follows: First, the spectrum of the current strain signal at the anchor point is extracted, and after Fourier transform, the difference between the spectrum and the average spectral profile obtained from historical data is calculated, and its absolute value is taken as the strain spectrum offset; next, the spectrum of the water pressure signal at the anchor point at the current moment is extracted, and the deviation between it and the average water pressure spectrum template under historical steady state is calculated. The absolute value is then multiplied by a frequency coupling factor to represent the relative influence of the water pressure signal in the overall judgment; finally, the spectral residual function is obtained by combining the strain spectrum offset and the relative influence of the water pressure signal in the overall judgment. Then, based on the spectral residual function, the residual energy density is calculated and compared with the preset energy density threshold. When the residual energy density is greater than the energy density threshold, it indicates that there is transient noise in the time period, and the phase remapping filter is called for noise reduction. The residual energy density is obtained as follows: First, the residual response function of the sensing anchor point in the frequency domain is obtained, and this residual function is integrated over a specified frequency range. To emphasize the strength of the deviation, the residual value at each frequency point is squared and then summed; finally, the residual energy density is obtained. The larger the residual energy density, the greater the difference between the signal at that point and the historical normal spectrum, indicating a potentially more significant hydraulic or structural anomaly. The normalization formula is as follows: ; In the formula, UWo represents the o-th data in the coupled dataset UW, YWo represents the o-th data in the original dataset YW, minYWo represents the valley value of the o-th data in the original dataset YW, and maxYWo represents the peak value of the o-th data in the original dataset YW.

[0008] Preferably, the signal decoupling and feature extraction module includes a water stress time-frequency decoupling unit and a time-domain cross-correlation and propagation feature extraction unit; The water stress time-frequency decoupling unit decouples the coupled dataset UW, extracts the axial micro-strain Ax and pore water pressure Pw of the anchor rod, and transforms them through short-time Fourier transform to obtain the time-frequency strain distribution function and the pore water pressure time-frequency distribution function; The time-frequency strain distribution function is obtained as follows: First, the axial micro-strain Ax sequence of the anchor rod collected by the sensing point on the time axis is extracted, and a time window is selected around time t. Within this time window, the strain signal of each instant is multiplied by a sine wave function containing different frequency components, and then these multiplications are accumulated and summed over the entire window. In this accumulation process, by integrating multiple times at different frequencies, the energy distribution of the axial micro-strain Ax of the anchor rod at each time period and each frequency is obtained, which is the time-frequency strain distribution function. The time-frequency distribution function of pore water pressure is obtained as follows: First, the system extracts the pore water pressure Pw collected by the sensing point on the time axis. Then, a local time window is selected in the signal sequence centered at time t. Within this window, the pore water pressure Pw at each time point is multiplied by a complex exponential function of a specific frequency. This complex exponential term acts as a "frequency scanner," separating different frequency components of the time signal. The product at each time point represents the instantaneous response of the signal at that frequency. Next, all these product results are integrated over the entire time window to obtain the time-frequency distribution function of pore water pressure. In other words, the frequency responses at each time point are summed up to obtain the overall energy and phase information of that frequency component at the current time t. When this operation is repeated at multiple different frequencies, the energy distribution of the signal at various times and frequencies can be obtained.

[0009] The obtained time-frequency strain distribution function and pore water pressure time-frequency distribution function are combined and normalized to obtain the independent water stress signal set Epw={nAx(i,t),nPw(i,t)}; In the formula, nAx(i,t) represents the decoupled axial strain signal of the i-th sensing anchor point at time t, and nPw(i,t) represents the decoupled pore water pressure signal of the i-th sensing anchor point at time t.

[0010] Preferably, the time-domain cross-correlation and propagation feature extraction unit is based on the independent water stress signal set Epw, calculates the correlation between water pressure disturbance and strain response, extracts the propagation features of hydraulic transient waves, and obtains the normalized cross-correlation function Cwp; The normalized cross-correlation function Cwp is obtained through the following formula: ; In the formula, tx represents the time delay variable, Cwp(i, tx) represents the normalized cross-correlation function of the i-th sensing anchor point under the condition of the time delay variable tx, d represents the integral sign, (t0, t1) represents the time integration interval, pnPw represents the average value of the decoupled pore water pressure signal, and pnAx represents the average value of the decoupled axial strain signal.

[0011] Preferably, the transient delay calculation and stress propagation direction matrix construction module includes a propagation delay calculation unit and a stress propagation direction matrix and vector intensity calculation unit; The propagation delay calculation unit searches for the peak point on the normalized cross-correlation function Cwp curve of each sensing point. When the normalized cross-correlation function Cwp(i, tx) of the i-th sensing anchor point reaches its maximum value under the condition of time delay variable tx, the time delay variable tx is marked and denoised to obtain the smooth time Δst. Collect all the acquired smoothing times Δst, and calculate the gradient field in the three-dimensional coordinate system through spatial difference to obtain the delayed gradient vector ∇Yst; The delay gradient vector ∇Yst is obtained using the following formula: ; In the formula, ∇ represents the gradient operator, ∂ represents the partial derivative symbol, ∂Δst / ∂x represents the time delay along the longitudinal direction of the tunnel, ∂Δst / ∂y represents the time delay along the horizontal extension direction of the surrounding rock, ∂Δst / ∂z represents the time delay along the vertical direction, and (x, y, z) represents the spatial coordinates.

[0012] Preferably, the stress propagation direction matrix and vector intensity calculation unit is based on the delayed gradient vector ∇Yst to construct the stress propagation direction matrix D and calculate the direction vector intensity Mdir; The stress propagation direction matrix D is obtained by the following formula: ; In the formula, D(x, y, z) represents the stress propagation direction matrix of coordinates (x, y, z); The direction vector strength Mdir is obtained using the following formula: ; In the formula, Mdir(i) represents the direction vector intensity of the i-th sensing anchor point, and (xi, yi, zi) represents the spatial coordinates of the i-th sensing anchor point.

[0013] Preferably, the collaborative response determination module includes a directional consistency assessment unit and a collaborative response index calculation unit; The orientation consistency evaluation unit constructs a spatial orientation vector field xU based on the stress propagation direction matrix D and the orientation vector intensity Mdir of each sensing anchor point; The spatial direction vector field xU is obtained by the following formula: xUi=-∇Yst(i) / Mdir(i); In the formula, xUi represents the spatial direction vector field of the i-th sensing anchor point, and ∇Yst(i) represents the time delay gradient vector of the i-th sensing anchor point; For each pair of adjacent sensing anchor points i and j, calculate the directional consistency coefficient Cang; The directional consistency coefficient (Cang) is obtained using the following formula: ; In the formula, Cang(i,j) represents the directional consistency coefficient between sensing anchor points i and j, and xUj represents the spatial direction vector field of the j-th sensing anchor point. The cooperative response index calculation unit combines the obtained directional consistency coefficient Cang with the decoupled axial strain signal nAx to calculate the cooperative response index Rc of the node. The synergistic response index Rc is obtained using the following formula: ; In the formula, Rc(i,j) represents the cooperative response index of sensing anchor points i and j, and nAx(j,t) represents the decoupled axial strain signal of the j-th sensing anchor point at time t.

[0014] Preferably, the tension control execution module includes an anchor bolt correction determination unit and a compensation tension force calculation and execution unit; The anchor bolt correction determination unit calculates the direction matching function SD between sensing anchor points i and j based on the stress propagation direction matrix D. The orientation matching function SD is obtained by dividing the dot product of the stress propagation direction matrix Di of the i-th sensing anchor point and the stress propagation direction matrix Dj of the j-th sensing anchor point by the magnitude of the stress propagation direction matrix Di of the i-th sensing anchor point and the stress propagation direction matrix Dj of the j-th sensing anchor point. The orientation matching degree function SD is combined with the cooperative response index Rc to calculate the orientation response joint decision function Ψadj, and compared with the preset decision threshold Ψth to obtain the anchor set Ωadj that needs to be corrected. The joint directional response determination function Ψadj is obtained through the following formula: Ψadj(i,j)=Rc(i,j)×SD(i,j); In the formula, Ψadj(i,j) represents the joint decision function of the direction response of sensing anchor points i and j, and SD(i,j) represents the direction matching degree function of sensing anchor points i and j; If sensing anchor points i and j satisfy this condition, then the marked node i enters the correction set, and the set of anchors that need to be corrected is Ωadj={i|∣|j,Ψadj(i,j)>Ψth}; The compensation tension calculation and execution unit corrects the sensing anchor points in the anchor bolt set Ωadj that needs to be corrected, and calculates the compensation tension ΔF; The compensating tension ΔF is obtained using the following formula: ; In the formula, ΔF(i) represents the required tension correction for the i-th sensing anchor point, and Rcavg(i) represents the average degree of cooperative imbalance of the i-th sensing anchor point. The average cooperative imbalance degree Rcavg(i) of the i-th sensing anchor point is obtained as follows: For the i-th sensing anchor point, firstly determine the adjacency set consisting of all its neighboring nodes, denoted as Ni. Each sensing anchor point j in this set has a propagation path with sensing anchor point i. Based on this, calculate the cooperative response index Rc(i,j) between sensing anchor point i and each of its neighboring sensing anchor points j. Then, sum up all the cooperative response indices one by one and divide by the total number of neighboring nodes Ni to obtain the result. Once the compensation tension ΔF is calculated, it is sent to the coordinated response determination module to re-analyze the coordinated response index Rc. When the average coordinated imbalance degree Rcavg(i) of the i-th sensing anchor point is less than the preset coordinated imbalance Rth, it indicates that the coordinated balance has been restored and the application is stopped; otherwise, the next adjustment cycle begins.

[0015] A tunnel hydraulic transient coupling support control method based on narrowband Internet of Things includes the following steps: Step 1: The data acquisition and processing module collects and preprocesses the tunnel hydraulic data through sensors installed between each anchor bolt and the surrounding rock to obtain the coupled dataset UW. Step 2: The signal decoupling and feature extraction module decouples the coupled dataset UW, extracts the propagation characteristics of the hydraulic transient wave, calculates the correlation between water pressure disturbance and strain response, and obtains the normalized cross-correlation function Cwp. Step 3: The transient delay calculation and stress propagation direction matrix construction module uses the normalized cross-correlation function Cwp as a basis to perform spatial interpolation of the anchor bolt's measurement points, calculate the spatial gradient, form the stress propagation direction matrix D, and obtain the direction vector intensity Mdir; Step 4: The collaborative response determination module uses the stress propagation direction matrix D and the direction vector intensity Mdir as a basis to perform a directional coupling determination function on the edge nodes and calculate the collaborative response index Rc of the nodes. Step 5: The tension control execution module combines the stress propagation direction matrix D and the cooperative response index Rc to obtain the set of anchor bolts Ωadj that need to be corrected, and calculates the compensation tension force ΔF.

[0016] This invention provides a tunnel hydraulic transient coupling support control system and method based on narrowband Internet of Things, which has the following beneficial effects: (1) During system operation, by incorporating fiber optic grating microstrain, pore water pressure signals, and spatial polar angles into a unified data acquisition framework, synchronous monitoring of a single anchor node across multiple physical dimensions is achieved. In particular, by acquiring the polar angle parameters in three-dimensional space through a miniature inertial measurement unit and a laser ranging module, each disturbance signal is equipped with a precise spatial positioning tag, breaking through the limitations of traditional monitoring which is limited to the time domain or scalar signal level, and providing a high-resolution spatial reference dimension for subsequent construction of disturbance direction and propagation path. When facing signal pollution problems in the complex environment of tunnels, traditional filtering methods often cannot cope with sudden noise or short-term anomalies.

[0017] (2) By extracting the maximum response peak value from the normalized cross-correlation function curve at each anchor point, the system can identify the actual response delay of water pressure disturbance propagation to that anchor point. This processing method no longer relies on the absolute signal amplitude or a set threshold, but instead focuses on the relative correlation time within the signal to construct a physically interpretable response hysteresis structure. Combined with noise removal and curve smoothing mechanisms, fluctuation interference caused by environmental disturbances and measurement point deviations can be eliminated, making the extracted propagation delay data have temporal continuity and repeatability, laying an accurate foundation for subsequent directional modeling.

[0018] By spatially differencing the propagation delay information extracted from different anchor points, the system constructs a delay gradient vector in a three-dimensional coordinate system. This vector not only reflects the speed of disturbance propagation from the source point to surrounding nodes but also describes the directionality of hydraulic wave penetration and diffraction in the surrounding rock medium. This spatial gradient modeling method based on time difference overcomes the instability of traditional stress path construction using strain or water pressure values, making the directionality of propagation trends clearer. It is particularly suitable for early identification of support status in irregular disturbance scenarios such as local unloading and shifting water inrush zones.

[0019] (3) By analyzing the direction vectors between adjacent anchor points, the system can not only detect whether there is a significant directional deviation, but also quantify the degree of deviation, providing a clear decision-making basis for whether to perform tension adjustment. This mechanism makes up for the blind spot in traditional support systems that rely solely on displacement or strain indicators to determine instability, and achieves accurate tracking from signal structure to directional field logic. In the tension control execution module, the system integrates the directional consistency coefficient with the stress propagation matrix of the anchor points to construct a joint directional response judgment function, thereby considering not only the response state of a single anchor point, but also evaluating its collaborative importance in group coupling.

[0020] (4) By deploying composite sensors at the interface between each anchor and the surrounding rock, synchronous sensing of pore water pressure and axial strain was achieved. In addition, a time alignment and normalization mechanism was introduced in the preprocessing process to construct a structure-water pressure joint coupling dataset, which enables the system to quickly locate the triggering position and propagation path of hydraulic transient waves under strong background disturbances, thus making up for the false alarm and missed alarm problems caused by traditional systems that rely on single-source data monitoring.

[0021] By constructing a three-dimensional stress propagation direction field through spatial difference analysis of transient propagation delay, and combining the direction vector intensity to determine the directionality and concentration of disturbance, the response consistency among the anchor bolt group is further analyzed using directional coupling functions. This process is no longer limited to judging the state of individual anchor bolts, but rather judges whether there is a region of coordinated imbalance based on the directional coordination among adjacent anchor bolts, providing a theoretical basis and triggering logic at the group behavior level for subsequent tension adjustment. Attached Figure Description

[0022] Figure 1 This is a schematic diagram of the block diagram of the tunnel hydraulic transient coupling support control system based on narrowband Internet of Things of the present invention; Figure 2 This is a schematic diagram of the steps of the tunnel hydraulic transient coupling support control method based on narrowband Internet of Things according to the present invention; Figure 3 This is a schematic diagram illustrating the process of obtaining the collaborative response index according to the present invention; Figure 4 This is a graph showing the trend of the synergistic response index of the present invention over time. Detailed Implementation

[0023] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0024] Example 1 This invention provides a tunnel hydraulic transient coupling support control system based on narrowband Internet of Things (IoT). Please refer to [link / reference]. Figures 1 to 4 It includes a data acquisition and processing module, a signal decoupling and feature extraction module, a transient delay calculation and stress propagation direction matrix construction module, a collaborative response determination module, and a tension control execution module; The data acquisition and processing module collects and preprocesses the tunnel hydraulic data through sensors installed between each anchor bolt and the surrounding rock to obtain the coupled dataset UW. The signal decoupling and feature extraction module decouples the coupled dataset UW, extracts the propagation characteristics of hydraulic transient waves, calculates the correlation between water pressure disturbance and strain response, and obtains the normalized cross-correlation function Cwp. The transient delay calculation and stress propagation direction matrix construction module is based on the normalized cross-correlation function Cwp. It performs spatial interpolation on the measurement points of the anchor bolt, calculates the spatial gradient, forms the stress propagation direction matrix D, and obtains the direction vector intensity Mdir. The collaborative response determination module uses the stress propagation direction matrix D and the direction vector intensity Mdir as a basis to perform a directional coupling determination function on the edge nodes and calculate the collaborative response index Rc of the nodes. The tension control execution module combines the stress propagation direction matrix D and the cooperative response index Rc to obtain the set of anchor bolts Ωadj that need to be corrected, and calculates the compensation tension force ΔF.

[0025] In this embodiment, by constructing a coupled dataset UW, the water pressure disturbance signal and the anchor bolt strain sequence are uniformly collected and processed. The correlation between them is identified through signal decoupling and feature extraction processes, and a dynamic linkage relationship between the transient disturbance source and the structural response is established, thereby providing the support system with dynamic sensing capabilities across physical fields.

[0026] This invention uses the normalized cross-correlation function as the core indicator to construct a transient propagation delay model, and forms a stress propagation direction matrix through spatial difference and gradient calculation of measurement points. This matrix not only reveals the spatial path of disturbance propagation, but also reflects the propagation concentration trend through the direction vector intensity. It can locate the potential unloading direction before the anchor bolt response occurs, realizing the shift from passive "anomaly detection" to active "trend early warning", and has the ability to predict geological emergencies in advance.

[0027] Traditional tunnel support systems often lack lateral coordination between structural units, especially in underground spaces with high impedance and easily distorted communication, making it difficult to efficiently share the response information of each anchor bolt. This invention broadcasts the stress propagation direction matrix to local anchor bolt groups via NB-IoT communication, enabling "edge nodes to share disturbance direction information." Based on this, a collaborative response index is determined to identify the consistency or deviation of stress response at the structural level, thereby guiding the adaptive tension control behavior among anchor bolts and constructing a truly meaningful swarm intelligence collaboration mechanism.

[0028] Conventional anchor bolt support systems, based on preset stress configurations, struggle to respond to heterogeneous evolution processes under environmental disturbances. This invention calculates the compensating tension force ΔF and introduces a directional correction function and a synergistic correction factor to specifically adjust the prestress state of key anchor bolts, achieving localized supplementation and directional control of support strength. This approach breaks free from the traditional constraint of "unadjustable initial settings," enabling the support system to possess self-evolution and self-repair capabilities during operation, significantly improving its service adaptability and lifecycle effectiveness.

[0029] Example 2 This embodiment is an explanation based on Embodiment 1. Please refer to it. Figure 1 and Figure 3 Specifically: the data acquisition and processing module includes a multi-source signal collaborative acquisition unit and a signal denoising and normalization fusion unit; The multi-source signal collaborative acquisition unit collects the axial micro-strain Ax of the anchor bolt, the pore water pressure Pw, and the spatial polar angle the of the sensing point through sensors installed between each anchor bolt and the surrounding rock; Among them, the axial micro-strain Ax of the anchor bolt is acquired by a fiber optic strain sensor array; The pore water pressure Pw is acquired by a pore water pressure sensor. The spatial polar angle the of the sensing point is acquired through a miniature inertial measurement unit and a laser ranging module; The original dataset YW was obtained by fitting the acquired axial micro-strain Ax of the anchor rod, pore water pressure Pw, and spatial polar angle the of the sensing point.

[0030] The signal denoising and normalization fusion unit preprocesses the original dataset YW to obtain the coupled dataset UW; Preprocessing includes data time alignment and synchronization, noise reduction, and normalization. Data time alignment and synchronization are performed by using an adaptive time remapping algorithm based on a local self-similar kernel function to align the data time of the original dataset YW. Denoising identifies anomalous noise segments by comparing residual spectral density. First, the spectral residual function of the signal is calculated: The spectral residual function is obtained as follows: First, the spectrum of the current strain signal at the anchor point is extracted. Then, the difference between the spectrum and the average spectral profile obtained from historical data is calculated, and its absolute value is taken as the strain spectrum offset. Next, the spectrum of the water pressure signal at the anchor point at the current moment is extracted, and the deviation between it and the average water pressure spectrum template under historical steady conditions is calculated. The absolute value is then multiplied by a frequency coupling factor to represent the relative influence of the water pressure signal in the overall judgment. Finally, the spectral residual function is obtained by combining the strain spectrum offset and the relative influence of the water pressure signal in the overall judgment. Then, based on the spectral residual function, the residual energy density is calculated and compared with the preset energy density threshold. When the residual energy density is greater than the energy density threshold, it indicates that there is transient noise in the time period, and the phase remapping filter is called for noise reduction. The normalization formula is as follows: ; In the formula, UWo represents the o-th data in the coupled dataset UW, YWo represents the o-th data in the original dataset YW, minYWo represents the valley value of the o-th data in the original dataset YW, and maxYWo represents the peak value of the o-th data in the original dataset YW.

[0031] In this embodiment, by incorporating fiber optic grating microstrain, pore water pressure signals, and spatial polar angles into a unified data acquisition framework, synchronous monitoring of a single anchor node across multiple physical dimensions is achieved. In particular, the acquisition of three-dimensional polar angle parameters using a miniature inertial measurement unit and a laser ranging module ensures that each disturbance signal possesses a precise spatial positioning tag. This overcomes the limitations of traditional monitoring, which is confined to the time domain or scalar signal level, and provides a high-resolution spatial reference dimension for subsequently constructing the disturbance direction and propagation path.

[0032] When facing signal contamination issues in the complex environment of tunnels, traditional filtering methods often fail to cope with sudden noise or short-term anomalies. This embodiment accurately identifies "abnormal segments" by comparing the residual of the historical average spectrum profile with the current spectrum, and achieves accurate detection of transient sudden noise by comparing the residual energy density with a threshold. Then, a phase remapping filter is applied to avoid the destruction of the original signal structure caused by conventional filtering, so that the disturbance mode still retains resolvability after noise reduction.

[0033] Example 3 This embodiment is an explanation based on Embodiment 2. Please refer to it. Figure 1 Specifically: the signal decoupling and feature extraction module includes a water stress time-frequency decoupling unit and a time-domain cross-correlation and propagation feature extraction unit; The water stress time-frequency decoupling unit decouples the coupled dataset UW, extracts the axial micro-strain Ax and pore water pressure Pw of the anchor rod, and transforms them through short-time Fourier transform to obtain the time-frequency strain distribution function and the pore water pressure time-frequency distribution function; The time-frequency strain distribution function is obtained as follows: First, the axial micro-strain Ax sequence of the anchor rod collected by the sensing point on the time axis is extracted, and a time window is selected around time t. Within this time window, the strain signal of each instant is multiplied by a sine wave function containing different frequency components, and then these multiplications are accumulated and summed over the entire window. In this accumulation process, by integrating multiple times at different frequencies, the energy distribution of the axial micro-strain Ax of the anchor rod at each time period and each frequency is obtained, which is the time-frequency strain distribution function. The time-frequency distribution function of pore water pressure is obtained as follows: First, the system extracts the pore water pressure Pw collected by the sensing point on the time axis. Then, a local time window is selected in the signal sequence with time t as the center. In this window, the pore water pressure Pw at each time point is multiplied by a complex exponential function of a specific frequency. Next, all these product results are integrated over the entire time window to obtain the time-frequency distribution function of pore water pressure. The obtained time-frequency strain distribution function and pore water pressure time-frequency distribution function are combined and normalized to obtain the independent water stress signal set Epw={nAx(i,t),nPw(i,t)}; In the formula, nAx(i,t) represents the decoupled axial strain signal of the i-th sensing anchor point at time t, and nPw(i,t) represents the decoupled pore water pressure signal of the i-th sensing anchor point at time t.

[0034] The temporal cross-correlation and propagation feature extraction unit is based on the independent water stress signal set Epw, calculates the correlation between water pressure disturbance and strain response, extracts the propagation features of hydraulic transient waves, and obtains the normalized cross-correlation function Cwp; The normalized cross-correlation function Cwp is obtained through the following formula: ; In the formula, tx represents the time delay variable, Cwp(i, tx) represents the normalized cross-correlation function of the i-th sensing anchor point under the condition of the time delay variable tx, d represents the integral sign, (t0, t1) represents the time integration interval, pnPw represents the average value of the decoupled pore water pressure signal, and pnAx represents the average value of the decoupled axial strain signal.

[0035] In this embodiment, by performing time-frequency deconstruction on the coupled dataset UW, short-time Fourier transforms are applied to the anchor bolt strain and pore water pressure respectively to obtain the frequency domain energy distribution at each moment. This achieves accurate mapping of the time-frequency distribution characteristics of stress and water pressure signals under dynamic processes. Compared with traditional single analysis methods based on the time or frequency domain, this approach can capture multi-scale fluctuation patterns in transient disturbances more meticulously. Especially under unstable states such as local unloading and water inrush in tunnels, it can effectively distinguish the anchor bolt response behavior caused by "background fluctuations" and "hydraulic pulses," thereby providing more discriminative feature inputs for accurate modeling of the propagation path.

[0036] In rock anchor systems, strain and water pressure often exhibit coupled responses, leading to blurred signal boundaries and redundant processing. This module addresses this by normalizing and reconstructing the time-frequency distribution function, independently mapping the strain and water pressure signals of each anchor point within each time period to form an independent set of water stress signals. This allows the system to perform subsequent correlation calculations while maintaining clear separation of physical quantities. This not only enhances the ability to identify the influence path of specific wave sources but also reduces false correlation judgments caused by signal coupling, making the construction of propagation delay and response relationships more physically consistent and traceable.

[0037] Based on the calculation of a normalized cross-correlation function using an independent set of water stress signals, the response hysteresis relationship between water pressure disturbance and anchor strain is effectively characterized. In this process, by introducing an integral window and an average offset correction, the actual delay of disturbance propagation between anchor points can be dynamically identified. Compared to traditional static correlation analysis, this normalized cross-correlation mechanism possesses time-varying properties, scale adaptability, and boundary sensitivity, enabling the extraction of "true delay characteristics" from numerous noise backgrounds, laying the foundation for subsequent construction of a stress propagation direction matrix and a collaborative response index.

[0038] Example 4 This embodiment is an explanation based on Embodiment 3. Please refer to it. Figure 3 Specifically: the transient delay calculation and stress propagation direction matrix construction module includes a propagation delay calculation unit and a stress propagation direction matrix and vector intensity calculation unit; The propagation delay calculation unit searches for the peak point on the normalized cross-correlation function Cwp curve of each sensing point. When the normalized cross-correlation function Cwp(i, tx) of the i-th sensing anchor point reaches its maximum value under the condition of time delay variable tx, the time delay variable tx is marked and denoised to obtain the smooth time Δst. Collect all the acquired smoothing times Δst, and calculate the gradient field in the three-dimensional coordinate system through spatial difference to obtain the delayed gradient vector ∇Yst; The delay gradient vector ∇Yst is obtained using the following formula: ; In the formula, ∇ represents the gradient operator, ∂ represents the partial derivative symbol, ∂Δst / ∂x represents the time delay along the longitudinal direction of the tunnel, ∂Δst / ∂y represents the time delay along the horizontal extension direction of the surrounding rock, ∂Δst / ∂z represents the time delay along the vertical direction, and (x, y, z) represents the spatial coordinates.

[0039] The stress propagation direction matrix and vector intensity calculation unit is based on the delayed gradient vector ∇Yst, constructs the stress propagation direction matrix D, and calculates and obtains the direction vector intensity Mdir; The stress propagation direction matrix D is obtained by the following formula: ; In the formula, D(x, y, z) represents the stress propagation direction matrix of coordinates (x, y, z); The direction vector strength Mdir is obtained using the following formula: ; In the formula, Mdir(i) represents the direction vector intensity of the i-th sensing anchor point, and (xi, yi, zi) represents the spatial coordinates of the i-th sensing anchor point.

[0040] In this embodiment, by extracting the maximum response peak value from the normalized cross-correlation function curve at each anchor point, the system can identify the actual response delay of water pressure disturbance propagation to that anchor point. This processing method no longer relies on absolute signal amplitude or set thresholds, but instead focuses on the relative correlation moments within the signal to construct a physically interpretable response hysteresis structure. Combined with noise removal and curve smoothing mechanisms, fluctuation interference caused by environmental disturbances and measurement point deviations can be eliminated, ensuring that the extracted propagation delay data has temporal continuity and repeatability, laying an accurate foundation for subsequent directional modeling.

[0041] By spatially differencing the propagation delay information extracted from different anchor points, the system constructs a delay gradient vector in a three-dimensional coordinate system. This vector not only reflects the speed of disturbance propagation from the source point to surrounding nodes but also describes the directionality of hydraulic wave penetration and diffraction in the surrounding rock medium. This spatial gradient modeling method based on time difference overcomes the instability of traditional stress path construction using strain or water pressure values, making the directionality of propagation trends clearer. It is particularly suitable for early identification of support status in irregular disturbance scenarios such as local unloading and shifting water inrush zones.

[0042] Based on the time-delay gradient vector, the system further constructs a stress propagation direction matrix, archives the disturbance direction trend at each coordinate point, and quantifies the actual impact level of water pressure disturbance on the anchor point by calculating the "direction vector intensity" index. This approach achieves a hierarchical mapping from "time-delay difference" to "disturbance directionality" and then to "response intensity," which can be used to identify potential high-response zones in the support system and serve as direct input for determining the adjustment target and direction of action in the subsequent tensioning control module, making the control behavior more targeted and aligned with the physical causal chain logic.

[0043] Example 5 This embodiment is an explanation based on Embodiment 4. Please refer to it. Figure 3 and Figure 4 Specifically: the collaborative response determination module includes a direction consistency assessment unit and a collaborative response index calculation unit; The orientation consistency evaluation unit constructs a spatial orientation vector field xU based on the stress propagation direction matrix D and the orientation vector intensity Mdir of each sensing anchor point; The spatial direction vector field xU is obtained by the following formula: xUi=-∇Yst(i) / Mdir(i); In the formula, xUi represents the spatial direction vector field of the i-th sensing anchor point, and ∇Yst(i) represents the time delay gradient vector of the i-th sensing anchor point; For each pair of adjacent sensing anchor points i and j, calculate the directional consistency coefficient Cang; The directional consistency coefficient (Cang) is obtained using the following formula: ; In the formula, Cang(i,j) represents the directional consistency coefficient between sensing anchor points i and j, and xUj represents the spatial direction vector field of the j-th sensing anchor point. The cooperative response index calculation unit combines the obtained directional consistency coefficient Cang with the decoupled axial strain signal nAx to calculate the cooperative response index Rc of the node. The synergistic response index Rc is obtained using the following formula: ; In the formula, Rc(i,j) represents the cooperative response index of sensing anchor points i and j, and nAx(j,t) represents the decoupled axial strain signal of the j-th sensing anchor point at time t.

[0044] The tension control execution module includes an anchor bolt correction judgment unit and a compensation tension force calculation and execution unit; The anchor bolt correction determination unit calculates the direction matching function SD between sensing anchor points i and j based on the stress propagation direction matrix D. The orientation matching function SD is obtained by dividing the dot product of the stress propagation direction matrix Di of the i-th sensing anchor point and the stress propagation direction matrix Dj of the j-th sensing anchor point by the magnitude of the stress propagation direction matrix Di of the i-th sensing anchor point and the stress propagation direction matrix Dj of the j-th sensing anchor point. The orientation matching degree function SD is combined with the cooperative response index Rc to calculate the orientation response joint decision function Ψadj, and compared with the preset decision threshold Ψth to obtain the anchor set Ωadj that needs to be corrected. The joint directional response determination function Ψadj is obtained through the following formula: Ψadj(i,j)=Rc(i,j)×SD(i,j); In the formula, Ψadj(i,j) represents the joint decision function of the direction response of sensing anchor points i and j, and SD(i,j) represents the direction matching degree function of sensing anchor points i and j; The set of anchors that need to be corrected is Ωadj={i|∣|j,Ψadj(i,j)>Ψth}; The compensation tension calculation and execution unit corrects the sensing anchor points in the anchor bolt set Ωadj that needs to be corrected, and calculates the compensation tension ΔF; The compensating tension ΔF is obtained using the following formula: ; In the formula, ΔF(i) represents the required tension correction for the i-th sensing anchor point, and Rcavg(i) represents the average degree of cooperative imbalance of the i-th sensing anchor point. The average cooperative imbalance degree Rcavg(i) of the i-th sensing anchor point is obtained as follows: For the i-th sensing anchor point, firstly determine the adjacency set consisting of all its neighboring nodes, denoted as Ni. Each sensing anchor point j in this set has a propagation path with sensing anchor point i. Based on this, calculate the cooperative response index Rc(i,j) between sensing anchor point i and each of its neighboring sensing anchor points j. Then, sum up all the cooperative response indices one by one and divide by the total number of neighboring nodes Ni to obtain the result. Once the compensation tension ΔF is calculated, it is sent to the coordinated response determination module to re-analyze the coordinated response index Rc. When the average coordinated imbalance degree Rcavg(i) of the i-th sensing anchor point is less than the preset coordinated imbalance Rth, it indicates that the coordinated balance has been restored and the application is stopped; otherwise, the next adjustment cycle begins.

[0045] In this embodiment, based on the stress propagation direction field and time-delay gradient information of each anchor point, a spatial direction consistency coefficient is introduced for the first time to characterize the similarity of the disturbance response trends between anchor points. By analyzing the direction vectors between adjacent anchor points, the system can not only detect whether there is a significant directional deviation, but also quantify the degree of deviation, providing a clear decision basis for whether to perform subsequent tensioning adjustments. This mechanism makes up for the blind spot in traditional support systems that rely solely on displacement or strain indicators to determine instability, achieving precise tracking from signal structure to direction field logic.

[0046] In the tension control execution module, the system integrates the directional consistency coefficient with the stress propagation matrix of the anchor points to construct a joint directional response judgment function. This not only considers the response state of a single anchor point but also assesses its collaborative importance in the group coupling. This approach effectively filters out false triggers caused by single-point fluctuations, giving the identified set of corrected anchors higher physical coupling logic credibility. Furthermore, after obtaining the target to be corrected, the system calculates its "average degree of collaborative imbalance" to further tailor the tension correction amount for each anchor point, avoiding over- or under-adjustment.

[0047] After each calculation of the compensation tension, the system does not directly declare the state complete. Instead, it feeds back the new applied state to the coordinated response judgment module for re-analysis to determine whether the system has returned to equilibrium. This approach breaks the static control logic of "one adjustment equals completion," forming a control loop of "identification → correction → re-identification → re-correction." More importantly, by setting a lower threshold for the degree of coordinated imbalance, the system possesses the ability to "self-stop," terminating operations promptly after equilibrium is restored. This ensures structural balance in the support force while avoiding ineffective energy consumption and unnecessary intervention.

[0048] Example 6 For a tunnel hydraulic transient coupling support control method based on narrowband Internet of Things, please refer to [reference needed]. Figure 2 Specifically, it includes the following steps: Step 1: The data acquisition and processing module collects and preprocesses the tunnel hydraulic data through sensors installed between each anchor bolt and the surrounding rock to obtain the coupled dataset UW. Step 2: The signal decoupling and feature extraction module decouples the coupled dataset UW, extracts the propagation characteristics of the hydraulic transient wave, calculates the correlation between water pressure disturbance and strain response, and obtains the normalized cross-correlation function Cwp. Step 3: The transient delay calculation and stress propagation direction matrix construction module uses the normalized cross-correlation function Cwp as a basis to perform spatial interpolation of the anchor bolt's measurement points, calculate the spatial gradient, form the stress propagation direction matrix D, and obtain the direction vector intensity Mdir; Step 4: The collaborative response determination module uses the stress propagation direction matrix D and the direction vector intensity Mdir as a basis to perform a directional coupling determination function on the edge nodes and calculate the collaborative response index Rc of the nodes. Step 5: The tension control execution module combines the stress propagation direction matrix D and the cooperative response index Rc to obtain the set of anchor bolts Ωadj that need to be corrected, and calculates the compensation tension force ΔF.

[0049] In this embodiment, by deploying composite sensors at the interface between each anchor bolt and the surrounding rock, synchronous sensing of pore water pressure and axial strain is achieved. Furthermore, a time-series alignment and normalization mechanism is introduced during preprocessing to construct a structure-water pressure joint coupling dataset. This enables the system to quickly locate the triggering position and propagation path of hydraulic transient waves under strong background disturbances, overcoming the false alarm and missed alarm problems caused by traditional systems relying solely on single-source data monitoring.

[0050] By constructing a three-dimensional stress propagation direction field through spatial difference analysis of transient propagation delay, and combining the direction vector intensity to determine the directionality and concentration of disturbance, the response consistency among the anchor bolt group is further analyzed using directional coupling functions. This process is no longer limited to judging the state of individual anchor bolts, but rather judges whether there is a region of coordinated imbalance based on the directional coordination among adjacent anchor bolts, providing a theoretical basis and triggering logic at the group behavior level for subsequent tension adjustment.

[0051] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A tunnel hydraulic transient coupling support control system based on narrowband Internet of Things, characterized in that: It includes a data acquisition and processing module, a signal decoupling and feature extraction module, a transient delay calculation and stress propagation direction matrix construction module, a coordinated response determination module, and a tension control execution module; The data acquisition and processing module collects and preprocesses the tunnel hydraulic data through sensors installed between each anchor bolt and the surrounding rock to obtain a coupled dataset. The signal decoupling and feature extraction module decouples the coupled dataset UW, extracts the propagation characteristics of hydraulic transient waves, calculates the correlation between water pressure disturbance and strain response, and obtains the normalized cross-correlation function. The transient delay calculation and stress propagation direction matrix construction module is based on the normalized cross-correlation function. It performs spatial interpolation on the measurement points of the anchor bolt, calculates the spatial gradient, forms the stress propagation direction matrix, and obtains the direction vector intensity. The collaborative response determination module uses the stress propagation direction matrix and direction vector intensity as a basis to perform a directional coupling determination function on the edge nodes and calculate the collaborative response index of the nodes. The tension control execution module combines the stress propagation direction matrix and the collaborative response index to obtain the set of anchor bolts that need to be corrected, and calculates the compensation tension force.

2. The tunnel hydraulic transient coupling support control system based on narrowband Internet of Things as described in claim 1, characterized in that: The data acquisition and processing module includes a multi-source signal collaborative acquisition unit and a signal denoising and normalization fusion unit; The multi-source signal collaborative acquisition unit collects the axial micro-strain Ax of the anchor bolt, the pore water pressure Pw, and the spatial polar angle the of the sensing point through sensors installed between each anchor bolt and the surrounding rock; Among them, the axial micro-strain Ax of the anchor bolt is acquired by a fiber optic strain sensor array; The pore water pressure Pw is acquired by a pore water pressure sensor. The spatial polar angle the of the sensing point is acquired through a miniature inertial measurement unit and a laser ranging module; The original dataset YW was obtained by fitting the acquired axial micro-strain Ax of the anchor rod, pore water pressure Pw, and spatial polar angle the of the sensing point.

3. The tunnel hydraulic transient coupling support control system based on narrowband Internet of Things as described in claim 2, characterized in that: The signal denoising and normalization fusion unit preprocesses the original dataset YW to obtain the coupled dataset UW; Preprocessing includes data time alignment and synchronization, noise reduction, and normalization. Data time alignment and synchronization are performed by using an adaptive time remapping algorithm based on a local self-similar kernel function to align the data time of the original dataset YW. Denoising identifies anomalous noise segments by comparing residual spectral density. First, the spectral residual function of the signal is calculated: The spectral residual function is obtained as follows: First, the spectrum of the current strain signal at the anchor point is extracted. Then, the difference between the spectrum and the average spectral profile obtained from historical data is calculated, and its absolute value is taken as the strain spectrum offset. Next, the spectrum of the water pressure signal at the anchor point at the current moment is extracted, and the deviation between it and the average water pressure spectrum template under historical steady conditions is calculated. The absolute value is then multiplied by a frequency coupling factor to represent the relative influence of the water pressure signal in the overall judgment. Finally, the spectral residual function is obtained by combining the strain spectrum offset and the relative influence of the water pressure signal in the overall judgment. Then, based on the spectral residual function, the residual energy density is calculated and compared with the preset energy density threshold. When the residual energy density is greater than the energy density threshold, it indicates that there is transient noise in the time period, and the phase remapping filter is called for noise reduction. The normalization formula is as follows: ； In the formula, UWo represents the o-th data in the coupled dataset UW, YWo represents the o-th data in the original dataset YW, minYWo represents the valley value of the o-th data in the original dataset YW, and maxYWo represents the peak value of the o-th data in the original dataset YW.

4. The tunnel hydraulic transient coupling support control system based on narrowband Internet of Things as described in claim 3, characterized in that: The signal decoupling and feature extraction module includes a water stress time-frequency decoupling unit and a time-domain cross-correlation and propagation feature extraction unit; The water stress time-frequency decoupling unit decouples the coupled dataset UW, extracts the axial micro-strain Ax and pore water pressure Pw of the anchor rod, and transforms them through short-time Fourier transform to obtain the time-frequency strain distribution function and the pore water pressure time-frequency distribution function; The time-frequency strain distribution function is obtained as follows: First, the axial micro-strain Ax sequence of the anchor rod collected by the sensing point on the time axis is extracted, and a time window is selected around time t. Within this time window, the strain signal of each instant is multiplied by a sine wave function containing different frequency components, and then these multiplications are accumulated and summed over the entire window. In this accumulation process, by integrating multiple times at different frequencies, the energy distribution of the axial micro-strain Ax of the anchor rod at each time period and each frequency is obtained, which is the time-frequency strain distribution function. The time-frequency distribution function of pore water pressure is obtained as follows: First, the system extracts the pore water pressure Pw collected by the sensing point on the time axis. Then, a local time window is selected in the signal sequence with time t as the center. In this window, the pore water pressure Pw at each time point is multiplied by a complex exponential function of a specific frequency. Next, all these product results are integrated over the entire time window to obtain the time-frequency distribution function of pore water pressure. The obtained time-frequency strain distribution function and pore water pressure time-frequency distribution function are combined and normalized to obtain the independent water stress signal set Epw={nAx(i,t),nPw(i,t)}; In the formula, nAx(i,t) represents the decoupled axial strain signal of the i-th sensing anchor point at time t, and nPw(i,t) represents the decoupled pore water pressure signal of the i-th sensing anchor point at time t.

5. The tunnel hydraulic transient coupling support control system based on narrowband Internet of Things as described in claim 4, characterized in that: The temporal cross-correlation and propagation feature extraction unit is based on the independent water stress signal set Epw, calculates the correlation between water pressure disturbance and strain response, extracts the propagation features of hydraulic transient waves, and obtains the normalized cross-correlation function Cwp; The normalized cross-correlation function Cwp is obtained through the following formula: ； In the formula, tx represents the time delay variable, Cwp(i, tx) represents the normalized cross-correlation function of the i-th sensing anchor point under the condition of the time delay variable tx, d represents the integral sign, (t0, t1) represents the time integration interval, pnPw represents the average value of the decoupled pore water pressure signal, and pnAx represents the average value of the decoupled axial strain signal.

6. The tunnel hydraulic transient coupling support control system based on narrowband Internet of Things as described in claim 5, characterized in that: The transient delay calculation and stress propagation direction matrix construction module includes a propagation delay calculation unit and a stress propagation direction matrix and vector intensity calculation unit; The propagation delay calculation unit searches for the peak point on the normalized cross-correlation function Cwp curve of each sensing point. When the normalized cross-correlation function Cwp(i, tx) of the i-th sensing anchor point reaches its maximum value under the condition of time delay variable tx, the time delay variable tx is marked and denoised to obtain the smooth time Δst. Collect all the acquired smoothing times Δst, and calculate the gradient field in the three-dimensional coordinate system through spatial difference to obtain the delayed gradient vector ∇Yst; The delay gradient vector ∇Yst is obtained using the following formula: ； In the formula, ∇ represents the gradient operator, ∂ represents the partial derivative symbol, ∂Δst / ∂x represents the time delay along the longitudinal direction of the tunnel, ∂Δst / ∂y represents the time delay along the horizontal extension direction of the surrounding rock, ∂Δst / ∂z represents the time delay along the vertical direction, and (x, y, z) represents the spatial coordinates.

7. The tunnel hydraulic transient coupling support control system based on narrowband Internet of Things as described in claim 6, characterized in that: The stress propagation direction matrix and vector intensity calculation unit is based on the delayed gradient vector ∇Yst, constructs the stress propagation direction matrix D, and calculates and obtains the direction vector intensity Mdir; The stress propagation direction matrix D is obtained by the following formula: ； In the formula, D(x, y, z) represents the stress propagation direction matrix of coordinates (x, y, z); The direction vector strength Mdir is obtained using the following formula: ； In the formula, Mdir(i) represents the direction vector intensity of the i-th sensing anchor point, and (xi, yi, zi) represents the spatial coordinates of the i-th sensing anchor point.

8. The tunnel hydraulic transient coupling support control system based on narrowband Internet of Things as described in claim 7, characterized in that: The collaborative response determination module includes a direction consistency assessment unit and a collaborative response index calculation unit; The orientation consistency evaluation unit constructs a spatial orientation vector field xU based on the stress propagation direction matrix D and the orientation vector intensity Mdir of each sensing anchor point; The spatial direction vector field xU is obtained by the following formula: xUi=-∇Yst(i) / Mdir(i); In the formula, xUi represents the spatial direction vector field of the i-th sensing anchor point, and ∇Yst(i) represents the time delay gradient vector of the i-th sensing anchor point; For each pair of adjacent sensing anchor points i and j, calculate the directional consistency coefficient Cang; The directional consistency coefficient (Cang) is obtained using the following formula: ; In the formula, Cang(i,j) represents the directional consistency coefficient between sensing anchor points i and j, and xUj represents the spatial direction vector field of the j-th sensing anchor point. The cooperative response index calculation unit combines the obtained directional consistency coefficient Cang with the decoupled axial strain signal nAx to calculate the cooperative response index Rc of the node. The synergistic response index Rc is obtained using the following formula: ； In the formula, Rc(i,j) represents the cooperative response index of sensing anchor points i and j, and nAx(j,t) represents the decoupled axial strain signal of the j-th sensing anchor point at time t.

9. The tunnel hydraulic transient coupling support control system based on narrowband Internet of Things as described in claim 8, characterized in that: The tension control execution module includes an anchor bolt correction judgment unit and a compensation tension force calculation and execution unit; The anchor bolt correction determination unit calculates the direction matching function SD between sensing anchor points i and j based on the stress propagation direction matrix D. The orientation matching function SD is obtained by dividing the dot product of the stress propagation direction matrix Di of the i-th sensing anchor point and the stress propagation direction matrix Dj of the j-th sensing anchor point by the magnitude of the stress propagation direction matrix Di of the i-th sensing anchor point and the stress propagation direction matrix Dj of the j-th sensing anchor point. The orientation matching degree function SD is combined with the cooperative response index Rc to calculate the orientation response joint decision function Ψadj, and compared with the preset decision threshold Ψth to obtain the anchor set Ωadj that needs to be corrected. The joint directional response determination function Ψadj is obtained through the following formula: Ψadj(i,j)=Rc(i,j)×SD(i,j); In the formula, Ψadj(i,j) represents the joint decision function of the direction response of sensing anchor points i and j, and SD(i,j) represents the direction matching degree function of sensing anchor points i and j; The set of anchors that need to be corrected is Ωadj={i|∣|j,Ψadj(i,j)>Ψth}; The compensation tension calculation and execution unit corrects the sensing anchor points in the anchor bolt set Ωadj that needs to be corrected, and calculates the compensation tension ΔF; The compensating tension ΔF is obtained using the following formula: ； In the formula, ΔF(i) represents the required tension correction for the i-th sensing anchor point, and Rcavg(i) represents the average degree of cooperative imbalance of the i-th sensing anchor point. The average cooperative imbalance degree Rcavg(i) of the i-th sensing anchor point is obtained as follows: For the i-th sensing anchor point, firstly determine the adjacency set consisting of all its neighboring nodes, denoted as Ni. Each sensing anchor point j in this set has a propagation path with sensing anchor point i. Based on this, calculate the cooperative response index Rc(i,j) between sensing anchor point i and each of its neighboring sensing anchor points j. Then, sum up all the cooperative response indices one by one and divide by the total number of neighboring nodes Ni to obtain the result. Once the compensation tension ΔF is calculated, it is sent to the coordinated response determination module to re-analyze the coordinated response index Rc. When the average coordinated imbalance degree Rcavg(i) of the i-th sensing anchor point is less than the preset coordinated imbalance Rth, it indicates that the coordinated balance has been restored and the application is stopped; otherwise, the next adjustment cycle begins.

10. A tunnel hydraulic transient coupling support control method based on narrowband Internet of Things, applied to the tunnel hydraulic transient coupling support control system based on narrowband Internet of Things as described in any one of claims 1 to 9, characterized in that: Includes the following steps: Step 1: The data acquisition and processing module collects and preprocesses the tunnel hydraulic data through sensors installed between each anchor bolt and the surrounding rock to obtain the coupled dataset UW. Step 2: The signal decoupling and feature extraction module decouples the coupled dataset UW, extracts the propagation characteristics of the hydraulic transient wave, calculates the correlation between water pressure disturbance and strain response, and obtains the normalized cross-correlation function Cwp. Step 3: The transient delay calculation and stress propagation direction matrix construction module uses the normalized cross-correlation function Cwp as a basis to perform spatial interpolation of the anchor bolt's measurement points, calculate the spatial gradient, form the stress propagation direction matrix D, and obtain the direction vector intensity Mdir; Step 4: The collaborative response determination module uses the stress propagation direction matrix D and the direction vector intensity Mdir as a basis to perform a directional coupling determination function on the edge nodes and calculate the collaborative response index Rc of the nodes. Step 5: The tension control execution module combines the stress propagation direction matrix D and the cooperative response index Rc to obtain the set of anchor bolts Ωadj that need to be corrected, and calculates the compensation tension force ΔF.

Images