Tunnel supporting structure stress wireless monitoring method and system

Abstract

The invention provides a tunnel supporting structure stress wireless monitoring method and system, and relates to the technical field of tunnel engineering monitoring, and the method comprises the steps: obtaining a stress response signal through a packaging housing, carrying out the time-frequency domain decomposition, constructing a space-time correlation diagram to recognize the topology of a stress transmission channel, determining a stress aggregation node, and constructing a local stress transmission coordinate system; deducing a high-risk area stress field, and constructing an event-driven wake-up mechanism according to a stress state transition moment to realize cooperative monitoring. According to the invention, a high-risk area can be accurately identified, energy consumption is reduced, and monitoring efficiency and early warning accuracy are improved.

Description

Technical Field

[0001] This invention relates to the field of tunnel engineering monitoring technology, and in particular to a wireless monitoring method and system for stress in tunnel support structures. Background Technology

[0002] Tunnel engineering is a crucial component of transportation infrastructure construction, and its safety directly impacts the safety of people's lives and property. As a key component ensuring tunnel stability, the stress state monitoring of the tunnel support structure is of paramount importance for tunnel safety. Traditional tunnel support structure monitoring primarily relies on wired sensing systems for data acquisition, using devices such as strain gauges and stress gauges to monitor stress changes in the support structure in real time. With the development of wireless sensing and Internet of Things (IoT) technologies, tunnel structure monitoring systems based on wireless sensor networks are increasingly being applied in engineering practice, enabling remote monitoring and real-time data transmission of tunnel support structures. Summary of the Invention

[0003] The present invention provides a method and system for wireless monitoring of stress in tunnel support structures, which can solve the problems in the prior art.

[0004] A first aspect of the present invention provides a method for wireless monitoring of stress in tunnel support structures, comprising:

[0005] Stress response signals are obtained by an encapsulation housing installed on the surface of the tunnel support structure. The encapsulation housing integrates a stress sensor, a microprocessor, a battery, and a wireless communication module.

[0006] On the microprocessor, the stress response signal is decomposed in the time and frequency domain to extract stress features characterizing the stress state of the support structure;

[0007] The stress characteristics are periodically uploaded to the monitoring center via a wireless communication module. Based on the stress characteristics, a spatiotemporal correlation diagram is constructed. The spatiotemporal correlation diagram represents the stress evolution synchronicity and spatial gradient relationship at the location of each encapsulation shell node. Based on the spatiotemporal correlation diagram, the dominant stress transmission channels are identified, and the stress transmission channel topology is obtained.

[0008] Based on the stress transmission channel topology, the stress convergence nodes of the support structure are determined and a local stress transmission coordinate system is constructed. Combined with adjacent encapsulation shell nodes, the stress distribution state around the stress convergence nodes is deduced to obtain the stress field in the high-risk area.

[0009] The stress state transition moment of the stress field in the high-risk area is identified, an event-driven wake-up mechanism for the encapsulation shell is constructed, a collaborative monitoring trigger command is calculated, and the collaborative monitoring trigger command is broadcast to the encapsulation shells around the stress convergence node through the wireless communication module. The awakened encapsulation shells enter the synchronous sampling mode and transmit the sampling data to the monitoring center in real time.

[0010] The stress response signal is decomposed in the time and frequency domain to extract stress features characterizing the stress state of the support structure, including:

[0011] The stress response signal is segmented into sliding windows, and the peak value, mean, and variance of the stress response signal are calculated within each time window to construct a time-domain statistical feature sequence.

[0012] The time-domain statistical feature sequence is subjected to a difference operation to calculate the peak increment, mean increment and variance increment between adjacent time windows to obtain the time-varying gradient feature. The time-varying gradient feature characterizes the dynamic change rate of the stress state of the support structure. The ratio of the peak increment to the variance increment is calculated to obtain the stress fluctuation concentration index.

[0013] The stress response signal is subjected to frequency domain transformation to extract the main frequency and the corresponding spectral amplitude, and a frequency domain feature vector is constructed. The stress fluctuation concentration index is concatenated with the frequency domain feature vector to form a time-frequency fusion feature.

[0014] The time-frequency fusion feature is subjected to principal component projection transformation, and the determinant value of the covariance matrix of multiple principal components is calculated to obtain the stress state stability quantification index. The determinant value characterizes the compactness of the distribution of the time-frequency fusion feature in the principal component space. The stress fluctuation concentration index, the time-frequency fusion feature and the stress state stability quantification index are combined as the stress feature output.

[0015] Based on the stress characteristics, a spatiotemporal correlation diagram is constructed. This diagram characterizes the synchronicity of stress evolution and the spatial gradient relationship at the locations of each encapsulation shell node, including:

[0016] The stress characteristic sequence of each package shell is periodically uploaded during continuous monitoring. The stress fluctuation concentration index is extracted and arranged in time order to form fluctuation time series data. The time-frequency fusion feature is extracted and arranged in time order to form time-frequency time series data.

[0017] The fluctuation time series data is reconstructed in phase space and mapped to a high-dimensional phase space using a time delay embedding method. Phase point trajectories are extracted in the high-dimensional phase space, and the maximum Lyapunov exponent of the phase point trajectories is calculated. The maximum Lyapunov exponents of any two encapsulation shells are extracted to calculate the peak value of the cross-correlation function, which is used as the dynamic evolution synchronization degree and an evolution synchronization correlation matrix is ​​constructed.

[0018] Singular value decomposition is performed on the time-frequency time-series data to extract the dominant singular value and calculate the time rate of change of the dominant singular value. The time rate of change difference between any two packaging shells is calculated. The spatial coordinates of each packaging shell are obtained and the spatial distance between the packaging shells is calculated. The time rate of change difference is divided by the spatial distance to obtain the spatial stress evolution gradient value and construct the spatial gradient correlation matrix.

[0019] Based on the quantification index of the stress state stability of each encapsulation shell, the evolution synchronization correlation matrix and the spatial gradient correlation matrix are constructed into a spatiotemporal coupling correlation matrix. The encapsulation shell is used as a node, and the element values ​​of the spatiotemporal coupling correlation matrix are used as edge weights to construct a spatiotemporal correlation graph.

[0020] Based on the spatiotemporal correlation diagram, the dominant stress transmission channels are identified, and the stress transmission channel topology is obtained, including:

[0021] The edge weights and spatial distances between adjacent encapsulation shell nodes are extracted from the spatiotemporal correlation graph. The stress correlation intensity per unit distance is calculated. The spatial stress evolution gradient values ​​between encapsulation shell nodes are extracted and vectorized in three-dimensional space to obtain the dominant direction representing the preferred stress transmission direction.

[0022] Based on the stress correlation intensity per unit distance and the dominant direction, multi-step diffusion with directional constraints is performed in the spatiotemporal correlation diagram. The product of the stress correlation intensity per unit distance on each side of the stress transmission path that meets the directional angle requirements is calculated to obtain the path stress transmission efficiency under directional constraints.

[0023] The absolute difference of the stress fluctuation concentration index between adjacent packaging shell nodes on the stress transmission path is calculated and accumulated to obtain the path stress fluctuation attenuation index. Combined with the path stress transmission efficiency value, the comprehensive transmission performance index of the path is calculated, and multiple stress transmission advantage channels are selected.

[0024] Extract the node identifiers, connection relationships, stress correlation intensity per unit distance, and comprehensive transmission performance indicators of the dominant direction and path of the stress transmission dominant channel, and output the stress transmission channel topology.

[0025] Based on the stress transmission channel topology, stress convergence nodes of the support structure are determined and a local stress transmission coordinate system is constructed. Combined with adjacent encapsulation shell nodes, the stress distribution around the stress convergence nodes is deduced, resulting in the stress field of the high-risk area, including:

[0026] Based on the stress transmission channel topology, the number of transmission paths entering each package shell node is counted, the variance of the angle between the dominant direction vectors of the entering paths is calculated to obtain the directional dispersion, and the product of the number of transmission paths entering is performed to obtain the stress convergence index. Package shell nodes with stress convergence index greater than the convergence threshold are regarded as stress convergence nodes.

[0027] The neighboring nodes of the stress convergence node are extracted, a local stress transfer coordinate system with the stress convergence node as the origin is established based on the dominant direction, the radial distance of the neighboring nodes is calculated, the stress correlation intensity per unit distance of the neighboring nodes is vector-projected along the dominant direction to obtain the effective stress transfer component, the stress attenuation gradient is obtained by quotienting it with the radial distance, and the stress attenuation field function is obtained by spatial interpolation fitting.

[0028] Based on the stress attenuation field function, a spatial monitoring grid is divided around the stress convergence node. The position parameters of each grid point in the local stress transfer coordinate system are calculated. Combined with the stress attenuation gradient of the neighboring nodes, the stress estimation value of the grid point is obtained. Grid points whose stress estimation value exceeds the stress risk threshold are extracted and connected and marked as high-risk area stress fields.

[0029] Based on the stress attenuation field function, a spatial monitoring grid is divided around the stress convergence node. The position parameters of each grid point in the local stress transfer coordinate system are calculated. Combined with the stress attenuation gradient of neighboring nodes, the estimated stress values ​​of the grid points are obtained, including:

[0030] Centered on the origin of the stress convergence node in the local stress transfer coordinate system, the outer boundary size of the spatial monitoring grid is determined according to the effective range of the stress attenuation field function. The spatial grid is divided into equal-interval sections along the three coordinate axes of the local stress transfer coordinate system to generate a set of grid nodes for the spatial monitoring grid and record their position parameters.

[0031] For each grid node, a subset of neighboring nodes whose spatial distance is less than the influence radius is extracted. The stress attenuation gradient of each neighboring node is multiplied by the inverse of its spatial distance to the grid node to obtain the distance-weighted stress attenuation gradient. The distance-weighted stress attenuation gradients of all neighboring nodes in the subset of neighboring nodes are summed and normalized to obtain the comprehensive stress attenuation gradient of the grid node.

[0032] Substitute the position parameters of the grid node into the stress attenuation field function to calculate the stress attenuation field function value at the grid node. Perform an inner product operation on the stress attenuation field function value and the comprehensive stress attenuation gradient in the dominant direction of the local stress transfer coordinate system, and superimpose the reference stress value at the stress convergence node to obtain the stress extrapolation value of the grid node.

[0033] Identify the stress state transition moments of the stress field in the high-risk area, construct an event-driven wake-up mechanism for the encapsulation shell, and calculate the collaborative monitoring trigger commands, including:

[0034] For the stress projection values ​​of each grid point in the stress field of the high-risk area, statistical features are extracted within a sliding window. The deviation between the current stress projection value and the historical average within the sliding window is calculated. When the deviation exceeds the abrupt change detection threshold, the current moment is marked as the stress state transition moment, and the spatial coordinates of the grid point where the abrupt change occurs are recorded.

[0035] Based on the stress inference values ​​of multiple historical moments before the stress state transition moment, a stress field spatiotemporal evolution trajectory model is constructed. The stress field spatiotemporal evolution trajectory model extrapolates and predicts the stress propagation direction and propagation speed after the stress state transition by fitting the motion trajectory of the stress inference values ​​in the time axis and spatial coordinate system, and generates a distribution map of the predicted arrival time of the stress wavefront.

[0036] The predicted arrival time distribution map of the stress wave front is superimposed and matched with the spatial position of the packaging shell. The packaging shell whose predicted arrival time is earlier than the current time plus a preset advance is identified as a pre-wake-up monitoring unit. The device identifier of the pre-wake-up monitoring unit and the difference between the predicted arrival time are encoded into a collaborative monitoring trigger command. The collaborative monitoring trigger command carries the wake-up countdown parameters of each pre-wake-up monitoring unit.

[0037] A second aspect of the present invention provides a wireless stress monitoring system for tunnel support structures, comprising:

[0038] The first unit is used to acquire stress response signals through an encapsulation shell installed on the surface of the tunnel support structure. The encapsulation shell integrates a stress sensor, a microprocessor, a battery, and a wireless communication module.

[0039] The second unit is used to perform time-frequency domain decomposition on the stress response signal on the microprocessor to extract stress features characterizing the stress state of the support structure.

[0040] The third unit is used to periodically upload the stress characteristics to the monitoring center through a wireless communication module, construct a spatiotemporal correlation diagram based on the stress characteristics, the spatiotemporal correlation diagram represents the stress evolution synchronicity and spatial gradient relationship at the location of each encapsulation shell node, and identify the dominant stress transmission channel based on the spatiotemporal correlation diagram to obtain the stress transmission channel topology.

[0041] The fourth unit is used to determine the stress convergence nodes of the support structure and construct a local stress transmission coordinate system based on the stress transmission channel topology. Combined with adjacent encapsulation shell nodes, it is used to deduce the stress distribution state around the stress convergence nodes and obtain the stress field of the high-risk area.

[0042] The fifth unit is used to identify the stress state transition time of the stress field in the high-risk area, construct an event-driven wake-up mechanism for the encapsulation shell, calculate the collaborative monitoring trigger command, and broadcast the collaborative monitoring trigger command to the encapsulation shells around the stress convergence node through the wireless communication module. The awakened encapsulation shells enter the synchronous sampling mode and transmit the sampling data to the monitoring center in real time.

[0043] A third aspect of the embodiments of the present invention,

[0044] An electronic device is provided, comprising:

[0045] processor;

[0046] Memory used to store processor-executable instructions;

[0047] The processor is configured to invoke instructions stored in the memory to execute the aforementioned method.

[0048] Fourth aspect of the present invention,

[0049] A computer-readable storage medium is provided, having stored thereon computer program instructions that, when executed by a processor, implement the aforementioned method.

[0050] The beneficial effects of this application are as follows:

[0051] By deploying an integrated housing containing stress sensors, microprocessors, batteries, and wireless communication modules on the surface of the tunnel support structure, distributed stress monitoring of the tunnel support structure is achieved, overcoming the shortcomings of traditional wired monitoring systems, such as difficult deployment and high maintenance costs.

[0052] The spatiotemporal correlation diagram constructed based on stress characteristics not only characterizes the synchronicity of stress evolution and spatial gradient relationship among monitoring nodes, but also identifies the dominant channels of stress transmission, thereby revealing the stress transmission mechanism in the tunnel support structure and providing a scientific basis for risk assessment.

[0053] An innovative event-driven wake-up mechanism based on the stress state transition moment was designed, which triggers collaborative monitoring only when an abnormal stress state is detected. This realizes a monitoring strategy that combines passive and active monitoring, which significantly extends battery life and reduces system maintenance costs while ensuring monitoring effectiveness. Attached Figure Description

[0054] Figure 1 This is a flowchart illustrating the wireless stress monitoring method for tunnel support structures according to an embodiment of the present invention.

[0055] Figure 2This is a flowchart illustrating the method for constructing a spatiotemporal correlation graph based on stress characteristics according to an embodiment of the present invention. Detailed Implementation

[0056] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0057] The technical solution of the present invention will be described in detail below with reference to specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments.

[0058] Figure 1 This is a flowchart illustrating the wireless stress monitoring method for tunnel support structures according to an embodiment of the present invention. Figure 1 As shown, the method includes:

[0059] Stress response signals are obtained by an encapsulation housing installed on the surface of the tunnel support structure. The encapsulation housing integrates a stress sensor, a microprocessor, a battery, and a wireless communication module.

[0060] On the microprocessor, the stress response signal is decomposed in the time and frequency domain to extract stress features characterizing the stress state of the support structure;

[0061] The stress characteristics are periodically uploaded to the monitoring center via a wireless communication module. Based on the stress characteristics, a spatiotemporal correlation diagram is constructed. The spatiotemporal correlation diagram represents the stress evolution synchronicity and spatial gradient relationship at the location of each encapsulation shell node. Based on the spatiotemporal correlation diagram, the dominant stress transmission channels are identified, and the stress transmission channel topology is obtained.

[0062] Based on the stress transmission channel topology, the stress convergence nodes of the support structure are determined and a local stress transmission coordinate system is constructed. Combined with adjacent encapsulation shell nodes, the stress distribution state around the stress convergence nodes is deduced to obtain the stress field in the high-risk area.

[0063] The stress state transition moment of the stress field in the high-risk area is identified, an event-driven wake-up mechanism for the encapsulation shell is constructed, a collaborative monitoring trigger command is calculated, and the collaborative monitoring trigger command is broadcast to the encapsulation shells around the stress convergence node through the wireless communication module. The awakened encapsulation shells enter the synchronous sampling mode and transmit the sampling data to the monitoring center in real time.

[0064] In one optional implementation, the stress response signal is decomposed in the time-frequency domain to extract stress features characterizing the stress state of the support structure, including:

[0065] The stress response signal is segmented into sliding windows, and the peak value, mean, and variance of the stress response signal are calculated within each time window to construct a time-domain statistical feature sequence.

[0066] The time-domain statistical feature sequence is subjected to a difference operation to calculate the peak increment, mean increment and variance increment between adjacent time windows to obtain the time-varying gradient feature. The time-varying gradient feature characterizes the dynamic change rate of the stress state of the support structure. The ratio of the peak increment to the variance increment is calculated to obtain the stress fluctuation concentration index.

[0067] The stress response signal is subjected to frequency domain transformation to extract the main frequency and the corresponding spectral amplitude, and a frequency domain feature vector is constructed. The stress fluctuation concentration index is concatenated with the frequency domain feature vector to form a time-frequency fusion feature.

[0068] The time-frequency fusion feature is subjected to principal component projection transformation, and the determinant value of the covariance matrix of multiple principal components is calculated to obtain the stress state stability quantification index. The determinant value characterizes the compactness of the distribution of the time-frequency fusion feature in the principal component space. The stress fluctuation concentration index, the time-frequency fusion feature and the stress state stability quantification index are combined as the stress feature output.

[0069] In this specific embodiment, it is necessary to obtain the stress response signal of the support structure. In practical applications, the stress response signal can be obtained by an encapsulation shell installed on the surface of the tunnel support structure. The encapsulation shell integrates a stress sensor, a microprocessor, a battery, and a wireless communication module to collect stress time series data over a period of time.

[0070] The acquired stress response signal is segmented into sliding windows. An appropriate time window size and sliding step size are set, for example, a window size of 60 seconds and a sliding step size of 10 seconds. Statistical characteristic values ​​are calculated for the data within each window. Within each time window, the peak value of the stress response signal is calculated, representing the maximum stress borne by the support structure within that window; the mean value of the stress response signal is calculated, reflecting the average stress level of the support structure within the window; and the variance of the stress response signal is calculated, characterizing the severity of stress fluctuations within the window. These three statistical quantities form a time-domain statistical characteristic sequence, recording the changing trend of the stress state of the support structure over time.

[0071] Differential operations are performed on the time-domain statistical feature sequence to calculate the feature changes between adjacent time windows. Specifically, for the i-th window and the (i-1)-th window, the peak increment, mean increment, and variance increment are calculated. These increments constitute the time-varying gradient feature, characterizing the dynamic rate of change of the stress state of the support structure. When the stress state of the support structure changes drastically, these increments will increase significantly. The ratio of the peak increment to the variance increment is calculated to obtain the stress fluctuation concentration index. This index reflects the relationship between the change in peak stress and the overall fluctuation change. When this value is large, it indicates that the stress concentration phenomenon is significant, predicting that the local area of ​​the support structure is subjected to excessive stress.

[0072] The original stress response signal is transformed in the frequency domain using the Fast Fourier Transform (FFT) method to convert the time-domain signal into a frequency-domain representation. In the frequency domain, the dominant frequency and its corresponding spectral amplitude are extracted to construct a frequency-domain feature vector. The dominant frequency reflects the main vibration characteristics of the support structure, while the spectral amplitude represents the intensity of the frequency component. The previously calculated stress fluctuation concentration index is concatenated with the frequency-domain feature vector to form a time-frequency fusion feature, which contains information from both the time and frequency domains.

[0073] Principal Component Analysis (PCA) projection transformation is performed on the time-frequency fusion features to standardize them and eliminate the influence of differences in the dimensions of different features. The covariance matrix of the features is calculated, and its eigenvalues ​​and eigenvectors are solved. The top few principal components with a cumulative contribution rate of 95% are selected, and the time-frequency fusion features are projected onto the principal component space. In the principal component space, the determinant values ​​of the covariance matrices of multiple principal components are calculated to obtain a quantitative index of stress state stability. This determinant value characterizes the compactness of the distribution of the time-frequency fusion features in the principal component space. The smaller the value, the more concentrated the feature distribution and the more stable the stress state of the support structure; the larger the value, the more dispersed the feature distribution and the more abnormal fluctuations in the stress state of the support structure.

[0074] The stress fluctuation concentration index, time-frequency fusion characteristics, and stress state stability quantification index are combined as stress feature outputs. These features comprehensively describe the stress state of the support structure, including static stress level, dynamic change trend, and stability assessment.

[0075] In practical applications, taking a tunnel project as an example, when the stress fluctuation concentration index at a monitoring point exceeds a preset threshold of 0.8, and the quantification index of stress state stability is greater than 0.5, it can be determined that the monitoring point is in an abnormal stress state and requires further inspection. This method enables real-time monitoring and early warning of the safety status of the support structure.

[0076] The aforementioned method extracts stress features that effectively characterize the stress state of the support structure through time-frequency domain signal decomposition and feature fusion, providing technical support for the safety monitoring and early warning of the support structure. This method can promptly detect abnormal stress on the support structure, effectively avoiding safety hazards caused by the instability of the support structure.

[0077] Figure 2 This is a flowchart illustrating the method for constructing a spatiotemporal correlation graph based on stress characteristics according to an embodiment of the present invention. In an optional implementation, a spatiotemporal correlation graph is constructed based on the stress characteristics. The spatiotemporal correlation graph characterizes the synchronicity of stress evolution and spatial gradient relationship at the locations of each encapsulation shell node, including:

[0078] The stress characteristic sequence of each package shell is periodically uploaded during continuous monitoring. The stress fluctuation concentration index is extracted and arranged in time order to form fluctuation time series data. The time-frequency fusion feature is extracted and arranged in time order to form time-frequency time series data.

[0079] The fluctuation time series data is reconstructed in phase space and mapped to a high-dimensional phase space using a time delay embedding method. Phase point trajectories are extracted in the high-dimensional phase space, and the maximum Lyapunov exponent of the phase point trajectories is calculated. The maximum Lyapunov exponents of any two encapsulation shells are extracted to calculate the peak value of the cross-correlation function, which is used as the dynamic evolution synchronization degree and an evolution synchronization correlation matrix is ​​constructed.

[0080] Singular value decomposition is performed on the time-frequency time-series data to extract the dominant singular value and calculate the time rate of change of the dominant singular value. The time rate of change difference between any two packaging shells is calculated. The spatial coordinates of each packaging shell are obtained and the spatial distance between the packaging shells is calculated. The time rate of change difference is divided by the spatial distance to obtain the spatial stress evolution gradient value and construct the spatial gradient correlation matrix.

[0081] Based on the quantification index of the stress state stability of each encapsulation shell, the evolution synchronization correlation matrix and the spatial gradient correlation matrix are constructed into a spatiotemporal coupling correlation matrix. The encapsulation shell is used as a node, and the element values ​​of the spatiotemporal coupling correlation matrix are used as edge weights to construct a spatiotemporal correlation graph.

[0082] In this specific embodiment, a spatiotemporal correlation diagram is constructed based on the stress characteristics. It is necessary to obtain the stress characteristic sequence periodically uploaded by each packaging shell during continuous monitoring. In practical applications, stress sensors can be set on the packaging shell to collect stress data of each node. For example, for the packaging shell of power equipment, stress sensor arrays can be arranged at key locations, with a sampling frequency set to 10 times per second, and continuous monitoring for 72 hours. Two types of key time series data can be extracted from the original stress data: fluctuation time series data and time-frequency time series data.

[0083] For the extraction of fluctuation time series data, a stress fluctuation concentration index is calculated. Specifically, a sliding window analysis is performed on the stress sequence of each packaging shell node, with a window length of 60 seconds and a step size of 5 seconds. Within each window, the ratio of the standard deviation to the mean of the stress value is calculated to obtain the fluctuation intensity, and the number of fluctuations exceeding a preset threshold is counted to obtain the fluctuation frequency. The product of the fluctuation intensity and the fluctuation frequency is used as the fluctuation concentration index for that time window. These indices are arranged in chronological order to form the fluctuation time series data.

[0084] For the extraction of time-frequency time-series data, wavelet transform was used to perform time-frequency analysis on the original stress data. Morlet wavelet was selected as the basic wavelet function, and the decomposition scale was set to 8 levels. Energy distribution characteristics at each scale were extracted, and the squared modulus of the wavelet coefficients was used as the time-frequency energy spectrum. Within each time window, statistical characteristics (mean, variance, kurtosis, skewness) and frequency band energy ratios of the energy spectrum were extracted and combined to form a time-frequency fusion feature vector. These feature vectors were arranged in chronological order to form the time-frequency time-series data.

[0085] The fluctuation time series data is reconstructed into a phase space and mapped to a high-dimensional phase space using a time-delay embedding method. Specifically, the optimal time delay τ is determined using mutual information; for example, for data with a sampling period of 0.1 seconds, τ is 7 sampling points. The embedding dimension m is determined using a pseudo-nearest neighbor algorithm, typically ranging from 4 to 8. For each encapsulation shell node i, the fluctuation time series data X... i (t), construct the trajectory points in the m-dimensional phase space: Y i (t)=[X i (t),X i (t+τ),X i (t+2τ),...,X i (t+(m-1)τ)].

[0086] In a high-dimensional phase space, the maximum Lyapunov exponent is calculated by analyzing the phase point trajectories. Using a small-data-quantity method, the nearest neighbor of each phase point is found, and the distance between them is tracked during subsequent evolution. If the initial distance is d0, the distance after k evolution steps is d... k Then we have d k ≈d0·eλ·k·dt, where λ is the maximum Lyapunov exponent and dt is the sampling time interval, obtained by applying ln(d k The λ value is obtained by linearly fitting / d0) with k·dt. To improve computational stability, this process is repeated multiple times (e.g., 100 times) and the average value is taken.

[0087] Calculate the cross-correlation function of the maximum Lyapunov exponents of any two package shells, and extract the peak value as the kinetic evolution synchronization degree. For the maximum Lyapunov exponent sequence λ of package shells i and j...i (t) and λ j (t), calculate the cross-correlation function R. ij (τ)=E[(λ i (t)-μ i )(λ j (t+τ)-μ j )] / σ i σ j Where μ and σ are the mean and standard deviation, respectively, and E is the expected value. Let R be the mean. ij The maximum value of (τ) is taken as the synchronization degree of dynamic evolution S. ij The synchronization degree between all the packaging shell pairs is formed into an n×n matrix S, which is the evolution synchronization correlation matrix.

[0088] For time-frequency and time-series data, singular value decomposition is performed to extract dominant dynamic features, and the time-frequency and time-series data of each encapsulation shell node i are organized into matrix M. i Where rows represent time windows and columns represent time-frequency feature dimensions, for M i Perform singular value decomposition: M i =U i ΣiV i T Where Σi is the singular value diagonal matrix, the maximum singular value σ is extracted. i1 As the dominant singular value, its rate of change over time r is calculated. i (t)=[σ i1 (t)-σ i1 (t-1)] / σ i1 (t-1).

[0089] Calculate the difference in the rate of change of time between any two package housings. For package housings i and j, calculate |r i (t)-r j The time average of (t)| yields the difference in rate of change D. ij Simultaneously, the spatial coordinates (x, y) of each package housing are obtained. i ,y i ,z i ), calculate the Euclidean distance L between the packaging shells. ij =sqrt[(x i -x j ) 2 +(y i -y j ) 2 +(z i -z j ) 2 Dividing the difference in the rate of change over time by the spatial distance yields the spatial stress evolution gradient value G. ij =D ij / Lij This forms an n×n spatial gradient correlation matrix G.

[0090] Based on the quantification index of the stress state stability of each package shell, the evolution synchronization correlation matrix and the spatial gradient correlation matrix are constructed into a spatiotemporal coupled correlation matrix. For each package shell i, its stress state stability index s is calculated. i For example, it can be characterized by being inversely proportional to the stress fluctuation amplitude or inversely proportional to the entropy value. Define the weighting factor w. i =s i / ∑s i Construct a spatiotemporal coupling correlation matrix C, whose elements are C0. ij =α·S ij +β·(1-G ij )·w i ·w j , where α and β are adjustable parameters that satisfy α+β=1, for example, α=0.6 and β=0.4.

[0091] Using the encapsulation shell as nodes and the elements of the spatiotemporal coupling correlation matrix as edge weights, a spatiotemporal correlation graph is constructed. In practical applications, a threshold θ (e.g., 0.7) can be set. When C ij When the value is greater than θ, a connection edge is established between nodes i and j, with an edge weight of C. ij By using the topology and edge weight distribution of the spatiotemporal correlation graph, the synchronicity of stress evolution and spatial gradient relationship between the nodes of the encapsulation shell can be intuitively reflected, providing an important basis for subsequent anomaly detection and fault diagnosis.

[0092] In one optional implementation, the dominant stress transmission channels are identified based on the spatiotemporal correlation diagram to obtain the stress transmission channel topology, including:

[0093] The edge weights and spatial distances between adjacent encapsulation shell nodes are extracted from the spatiotemporal correlation graph. The stress correlation intensity per unit distance is calculated. The spatial stress evolution gradient values ​​between encapsulation shell nodes are extracted and vectorized in three-dimensional space to obtain the dominant direction representing the preferred stress transmission direction.

[0094] Based on the stress correlation intensity per unit distance and the dominant direction, multi-step diffusion with directional constraints is performed in the spatiotemporal correlation diagram. The product of the stress correlation intensity per unit distance on each side of the stress transmission path that meets the directional angle requirements is calculated to obtain the path stress transmission efficiency under directional constraints.

[0095] The absolute difference of the stress fluctuation concentration index between adjacent packaging shell nodes on the stress transmission path is calculated and accumulated to obtain the path stress fluctuation attenuation index. Combined with the path stress transmission efficiency value, the comprehensive transmission performance index of the path is calculated, and multiple stress transmission advantage channels are selected.

[0096] Extract the node identifiers, connection relationships, stress correlation intensity per unit distance, and comprehensive transmission performance indicators of the dominant direction and path of the stress transmission dominant channel, and output the stress transmission channel topology.

[0097] In this specific embodiment, edge weights and spatial distance data between adjacent encapsulation shell nodes are extracted from the spatiotemporal correlation graph. The edge weights are obtained by calculating the correlation coefficient of the stress time series between nodes, and the spatial distance is determined by measuring the actual physical distance between adjacent sensor nodes. For example, for two adjacent nodes A and B at the tunnel arch, the edge weight is 0.85 and the spatial distance is 2.5 meters.

[0098] To eliminate the influence of distance differences between different nodes on stress correlation, the stress correlation strength per unit distance is calculated. The specific method is to divide the edge weight by the corresponding spatial distance. Taking the aforementioned nodes A and B as examples, the stress correlation strength per unit distance is 0.85 / 2.5=0.34. This index can more objectively reflect the stress transmission efficiency in a unit space, making the stress correlation strength between nodes with different distances comparable.

[0099] Extract the spatial stress evolution gradient values ​​between the nodes of the encapsulation shell, calculate the difference in stress change between adjacent nodes within a specific time window, and then divide the difference by the physical distance between the nodes to obtain the stress gradient. Perform vector decomposition on this gradient in three-dimensional space to obtain the components in the X, Y, and Z directions. By comparing the magnitudes of these three components, determine the dominant direction of stress transmission. For example, if the stress gradient component of a certain pair of adjacent nodes is the largest in the X direction, then the X direction is the dominant direction of stress transmission.

[0100] Based on the calculated stress correlation intensity per unit distance and the dominant direction, multi-step diffusion of directional constraints is performed in the spatiotemporal correlation graph. A directional angle threshold is set. For example, a path with an angle of no more than 30 degrees to the dominant direction is considered to satisfy the directional constraint condition. Starting from the starting node, the diffusion proceeds step by step along the edges that satisfy the directional constraints. The product of the stress correlation intensity per unit distance on each edge of the path is calculated as the stress transfer efficiency of the path. For the path extending from the arch to both sides in the tunnel support structure, there are multiple paths that satisfy the directional constraints, and each path has its corresponding stress transfer efficiency.

[0101] The absolute differences in stress fluctuation concentration indices between adjacent packaging shell nodes along the stress transmission path are calculated and summed to obtain the path stress fluctuation attenuation index. The stress fluctuation concentration index is calculated based on the fluctuation characteristics of the stress time series at the nodes, reflecting the degree of attenuation of the stress signal during transmission. For example, for a path containing nodes A, B, and C, the stress fluctuation attenuation index for that path is obtained by calculating |fluctuation index A - fluctuation index B| + |fluctuation index B - fluctuation index C|.

[0102] The comprehensive stress transfer performance index of the path can be calculated by combining the path stress transfer efficiency and the path stress fluctuation attenuation index. A weighted method can be used: Comprehensive stress transfer performance index of the path = α × path stress transfer efficiency - β × path stress fluctuation attenuation index, where α and β are weighting coefficients determined according to actual monitoring needs. In stress monitoring of tunnel support structures, the weights can be adjusted according to the material characteristics and engineering requirements of the support structure. For example, for rigid support structures, the value of α can be increased to focus more on stress transfer efficiency.

[0103] By setting a threshold for the comprehensive stress transmission performance index, multiple stress transmission dominant channels are selected. These channels are the paths through which stress is preferentially transmitted in the tunnel support structure. They are of great significance for monitoring the stress distribution of the support structure and predicting potential risk areas. For example, if the stress value on a certain stress transmission dominant channel suddenly increases during tunnel construction, it indicates that the support structure is about to face the risk of instability in that area.

[0104] Extract the node identifiers, connection relationships, stress correlation intensity per unit distance, dominant direction, and comprehensive transmission performance indicators of the dominant stress transmission channels, and output the stress transmission channel topology. This topology visually displays the dominant stress transmission paths in a graphical manner. Nodes represent the location of the encapsulation shell, edges represent stress transmission relationships, edge thickness indicates the magnitude of stress correlation intensity, and edge direction indicates the stress transmission direction. Color coding represents the comprehensive transmission performance indicators of different paths, facilitating engineers to quickly identify key monitoring areas.

[0105] In practical applications, this method can be used for stress monitoring of support structures under complex geological conditions such as tunnels crossing fault zones. For example, in a section of a high-speed railway tunnel crossing a fault zone, a monitoring network was formed by arranging multiple encapsulated shells to collect stress data of each node of the support structure in real time. The stress transmission dominance channel identified by this method showed that stress was mainly transmitted from near the fault zone to both sides of the tunnel, and the stress transmission efficiency of the left support structure was significantly higher than that of the right. Based on this finding, the engineers increased the monitoring frequency of the left support structure and adjusted the support parameters in a timely manner, thus avoiding the risk of instability of the support structure.

[0106] In one optional implementation, based on the stress transmission channel topology, stress convergence nodes of the support structure are determined and a local stress transmission coordinate system is constructed. Combined with adjacent encapsulation shell nodes, the stress distribution around the stress convergence nodes is deduced to obtain the stress field in the high-risk area, including:

[0107] Based on the stress transmission channel topology, the number of transmission paths entering each package shell node is counted, the variance of the angle between the dominant direction vectors of the entering paths is calculated to obtain the directional dispersion, and the product of the number of transmission paths entering is performed to obtain the stress convergence index. Package shell nodes with stress convergence index greater than the convergence threshold are regarded as stress convergence nodes.

[0108] The neighboring nodes of the stress convergence node are extracted, a local stress transfer coordinate system with the stress convergence node as the origin is established based on the dominant direction, the radial distance of the neighboring nodes is calculated, the stress correlation intensity per unit distance of the neighboring nodes is vector-projected along the dominant direction to obtain the effective stress transfer component, the stress attenuation gradient is obtained by quotienting it with the radial distance, and the stress attenuation field function is obtained by spatial interpolation fitting.

[0109] Based on the stress attenuation field function, a spatial monitoring grid is divided around the stress convergence node. The position parameters of each grid point in the local stress transfer coordinate system are calculated. Combined with the stress attenuation gradient of the neighboring nodes, the stress estimation value of the grid point is obtained. Grid points whose stress estimation value exceeds the stress risk threshold are extracted and connected and marked as high-risk area stress fields.

[0110] In this specific embodiment, stress transmission channel topology data is obtained. This data includes information on the packaging shell nodes and their stress transmission relationships. For each packaging shell node, the number N of its transmission paths is counted. For example, if a node A receives stress transmission from nodes B, C, and D, then the number of transmission paths of node A is 3.

[0111] Calculate the variance of the angle between the dominant direction vectors of the merging path. For node A, extract the direction vectors from all nodes connected to it to node A, denoted as v1, v2, ..., v1. n These direction vectors are normalized to make them all unit vectors. The variance of the angles between these unit vectors is calculated and denoted as σ. 2 The larger the variance, the more dispersed the stress transmission direction; the smaller the variance, the more concentrated the stress transmission direction.

[0112] Calculate the stress convergence index I, and incorporate the number N and directional dispersion σ of the transmission paths. 2 Multiplication, i.e., I = N × σ 2 When I is greater than the preset convergence threshold T, the node is marked as a stress convergence node. The convergence threshold T can be adjusted according to actual engineering needs, and is usually in the range of 5 to 15. In a practical case, assuming the convergence threshold T is set to 10, if the number of transmission paths merging into a node is 5 and the directional dispersion is 2.5, then the stress convergence index is 12.5, which is greater than the threshold of 10, and the node is identified as a stress convergence node.

[0113] After identifying the stress convergence nodes, extract their neighborhood node set. The neighborhood can be defined as all nodes with a Euclidean distance not exceeding d, where d is typically taken as 10% to 20% of the characteristic dimension of the support structure. For each stress convergence node, calculate its dominant direction vector v̄. The dominant direction vector can be obtained by weighted averaging of the direction vectors of the merging paths, with the weights set to the stress transfer intensity of the corresponding path.

[0114] A local stress transfer coordinate system is established based on the dominant direction vector v̄ and the location of stress convergence nodes. With the stress convergence node as the origin O and the direction of the dominant direction vector v̄ as the positive x-axis, vectors perpendicular to v̄ are selected as the y-axis and z-axis, forming a right-handed coordinate system. In this coordinate system, the radial distance r of neighboring nodes, i.e., the distance from the neighboring node to the origin, is calculated.

[0115] For each neighboring node, the stress correlation intensity S per unit distance can be calculated by dividing the stress transfer coefficient between nodes by the distance between nodes. The effective stress transfer component S is then obtained by vector projection of S along the dominant direction v̄. eff =S·cosθ, where θ is the angle between the line connecting the nodes and the dominant direction. Let S... eff The stress attenuation gradient G = S is obtained by quoting the radial distance r. eff / r.

[0116] Using the stress attenuation gradient values ​​of multiple neighboring nodes, the stress attenuation field function f(x,y,z) is fitted using a spatial interpolation method. In practical applications, radial basis functions can be used for interpolation, with the form f(x,y,z)=Σ(w i ·φ(||pp i ||)), where p=(x,y,z) are the coordinates of any point in space, p i Given the coordinates of the neighboring nodes, w i φ is the weighting coefficient, and φ is the radial basis function.

[0117] Based on the stress attenuation field function f(x,y,z), a spatial monitoring grid is divided around the stress convergence node. The grid size can be set to 1%~5% of the characteristic size of the support structure. For each grid point, its position parameters (x,y,z) in the local stress transfer coordinate system are calculated and substituted into the stress attenuation field function. Combined with the stress attenuation gradient of the neighboring nodes, the estimated stress value S of that grid point is obtained. pred =f(x,y,z).

[0118] Set stress risk threshold S threshold This is typically taken as 80% to 90% of the material's allowable stress. Extracting stress values ​​exceeding S... thresholdThe grid points are used to form a set of high-risk points. Through regional connectivity analysis, adjacent high-risk points are connected and marked to form a stress field in the high-risk region.

[0119] In a practical application of a concrete underground tunnel support structure, the above method was used to identify three main stress convergence nodes, and a high-risk stress field with an area of ​​0.25m² was constructed around these nodes. 2 0.18m 2 and 0.31m 2 By strengthening the support measures in these areas, the overall safety factor of the tunnel structure was improved, verifying the effectiveness of the method.

[0120] In one optional implementation, based on the stress attenuation field function, a spatial monitoring grid is divided around the stress convergence node. The position parameters of each grid point in the local stress transfer coordinate system are calculated. Combined with the stress attenuation gradient of neighboring nodes, the stress estimation value of the grid point is obtained, including:

[0121] Centered on the origin of the stress convergence node in the local stress transfer coordinate system, the outer boundary size of the spatial monitoring grid is determined according to the effective range of the stress attenuation field function. The spatial grid is divided into equal-interval sections along the three coordinate axes of the local stress transfer coordinate system to generate a set of grid nodes for the spatial monitoring grid and record their position parameters.

[0122] For each grid node, a subset of neighboring nodes whose spatial distance is less than the influence radius is extracted. The stress attenuation gradient of each neighboring node is multiplied by the inverse of its spatial distance to the grid node to obtain the distance-weighted stress attenuation gradient. The distance-weighted stress attenuation gradients of all neighboring nodes in the subset of neighboring nodes are summed and normalized to obtain the comprehensive stress attenuation gradient of the grid node.

[0123] Substitute the position parameters of the grid node into the stress attenuation field function to calculate the stress attenuation field function value at the grid node. Perform an inner product operation on the stress attenuation field function value and the comprehensive stress attenuation gradient in the dominant direction of the local stress transfer coordinate system, and superimpose the reference stress value at the stress convergence node to obtain the stress extrapolation value of the grid node.

[0124] In this specific embodiment, the outer boundary size of the spatial monitoring grid is determined with the origin of the stress convergence node in the local stress transfer coordinate system as the center. Specifically, it is determined based on the effective range R of the stress attenuation field function. eff The side length of the spatial monitoring cube is set to 2×R. eff To ensure coverage of the entire area affected by stress, for example, if the effective range of the stress attenuation field function is 100 meters, then a cubic spatial monitoring area with a side length of 200 meters is established with the stress convergence node as the center.

[0125] The local stress transfer coordinate system is spatially partitioned at equal intervals along the X, Y, and Z axes. The partitioning accuracy can be determined according to actual engineering requirements; for example, 20 cells can be evenly divided in each direction to form 21×21×21 grid nodes. For each grid node, its three-dimensional coordinate position (x, y, z) in the local stress transfer coordinate system is recorded. i ,y i ,z i ) are used as position parameters, and these nodes are organized into a set of grid nodes G.

[0126] For each grid node g in the set of grid nodes G i Extracting spatial distances smaller than the influence radius r inf The neighborhood node subset N i In practice, this can be achieved by calculating the Euclidean distances from all known stress value nodes around the stress convergence node to the current mesh node, and then filtering out nodes with distances less than r. inf The node. Assume the influence radius r inf If set to 50 meters, then for grid node g i Calculate all known stress values ​​at nodes to g_ i Nodes within a distance of less than 50 meters are selected to form a neighborhood node subset N. i .

[0127] For the neighborhood node subset N i Each neighbor node n in j The stress attenuation gradient is calculated by taking the partial derivative of the stress attenuation field function with respect to spatial coordinates, and is expressed as a vector grad. j At the same time, calculate the neighboring nodes n j To grid node g i Spatial distance d j The stress attenuation gradient grad j and the reciprocal of the distance 1 / d j Multiplying these yields the distance-weighted stress attenuation gradient w. gradj =grad j ×(1 / d j This approach ensures that nodes that are closer together have a greater impact on the stress of the mesh points, and that the impact is greater on the neighboring node subset N. i The distance-weighted stress attenuation gradients of all nodes are summed, and then normalized by dividing the sum by the sum of all weights to obtain the mesh node g. i The combined stress attenuation gradient G gradi :

[0128] G gradi =∑(w gradj) / ∑(1 / d j ),

[0129] grid node g i Position parameter (x) i ,y i ,z i Substituting the stress attenuation field function F, the stress attenuation field function value F(x) is calculated. i ,y i ,z i The stress decay field function is typically a function that decays with distance, for example, based on an exponential decay model:

[0130] F(x,y,z)=exp(-√(x 2 +y 2 +z 2 ) / α), where α is the attenuation characteristic constant, and the stress attenuation field function value F(x) is... i ,y i ,z i ) and the combined stress attenuation gradient G gradi Perform the inner product operation along the dominant direction v̄ of the local stress transfer coordinate system:

[0131] stress contribution =F(x i ,y i ,z i )·(G gradi ·v̄),

[0132] The dominant direction v̄ represents the predominant direction of stress transfer. The above calculation results are compared with the reference stress value at the stress convergence node. base Superimposed, we obtain the mesh node g. i The estimated stress value i :

[0133] stress i =stress base +stress contribution，

[0134] Through the above calculation process, the stress projection values ​​of each grid point in the spatial monitoring grid can be obtained, thereby constructing a complete stress distribution field. This method is particularly suitable for scenarios in geotechnical engineering such as deep mining and tunnel construction that require assessment of the stress distribution of the surrounding rock mass. For example, in deep mining operations, the stress distribution of the entire mining area can be projected based on the stress measurement values ​​of several key measuring points, providing a basis for support design and safe production.

[0135] In practical applications, parameters such as the radius of influence and mesh density can be adjusted according to specific engineering needs to balance calculation accuracy and efficiency. Furthermore, stress simulation results can be presented using 3D visualization technology, intuitively displaying stress distribution patterns and assisting in engineering decision-making.

[0136] In one optional implementation, the stress state transition moment of the stress field in the high-risk area is identified, an event-driven wake-up mechanism for the encapsulation shell is constructed, and a collaborative monitoring trigger command is calculated, including:

[0137] For the stress projection values ​​of each grid point in the stress field of the high-risk area, statistical features are extracted within a sliding window. The deviation between the current stress projection value and the historical average within the sliding window is calculated. When the deviation exceeds the abrupt change detection threshold, the current moment is marked as the stress state transition moment, and the spatial coordinates of the grid point where the abrupt change occurs are recorded.

[0138] Based on the stress inference values ​​of multiple historical moments before the stress state transition moment, a stress field spatiotemporal evolution trajectory model is constructed. The stress field spatiotemporal evolution trajectory model extrapolates and predicts the stress propagation direction and propagation speed after the stress state transition by fitting the motion trajectory of the stress inference values ​​in the time axis and spatial coordinate system, and generates a distribution map of the predicted arrival time of the stress wavefront.

[0139] The predicted arrival time distribution map of the stress wave front is superimposed and matched with the spatial position of the packaging shell. The packaging shell whose predicted arrival time is earlier than the current time plus a preset advance is identified as a pre-wake-up monitoring unit. The device identifier of the pre-wake-up monitoring unit and the difference between the predicted arrival time are encoded into a collaborative monitoring trigger command. The collaborative monitoring trigger command carries the wake-up countdown parameters of each pre-wake-up monitoring unit.

[0140] In this specific embodiment, statistical features are extracted within a sliding window for the estimated stress values ​​of each grid point in the stress field of the high-risk area. The estimated stress values ​​are the stress state estimates of each grid point calculated based on measured stress data and stress propagation models. The sliding window length is set to T time units, and the window contains the estimated stress values ​​of the most recent T time points. Statistical features are extracted within this sliding window, including features such as mean, standard deviation, peak value, and valley value. For example, for a grid point on the tunnel arch, the time series data of its estimated stress values ​​over the past 24 hours are processed, and the mean stress within the sliding window is calculated to be 15 MPa, and the standard deviation is 2.5 MPa.

[0141] The deviation between the current estimated stress value and the historical mean within the sliding window is calculated. The deviation can be obtained by dividing the difference between the current stress value and the historical mean by the historical standard deviation, which indicates the degree to which the current stress deviates from the normal state. A sudden change detection threshold α is set. When the deviation exceeds α, the current moment is marked as the stress state transition moment, and the spatial coordinates of the grid point where the sudden change occurs are recorded. Taking a grid point on the tunnel arch as an example, if the current estimated stress value is 22 MPa, the historical mean is 15 MPa, and the standard deviation is 2.5 MPa, then the deviation is (22-15) / 2.5=2.8. If the sudden change detection threshold α is set to 2.5, then the current moment is determined to be the stress state transition moment.

[0142] Based on the inferred stress values ​​from multiple historical moments prior to the stress state transition, a spatiotemporal evolution trajectory model of the stress field is constructed. This model is achieved by fitting the motion trajectory of the inferred stress values ​​in the time axis and spatial coordinate system. Specifically, the stress value changes of the spatial grid around the abrupt change point at different moments are collected, the spatial direction and velocity characteristics of stress propagation are analyzed, and regression analysis, trend extrapolation, and other methods are applied to establish a predictive model for stress propagation.

[0143] The spatiotemporal evolution trajectory model of the stress field considers the material properties of the support structure, the propagation law of stress waves, and the propagation patterns observed in historical data. For example, for concrete support structures, the propagation speed of stress waves is related to parameters such as the elastic modulus and density of the material. Through model fitting, the propagation direction vector and propagation speed scalar of stress waves in three-dimensional space are obtained. For example, in an elastic surrounding rock section of a tunnel, the propagation speed of stress waves is about 3000 meters per second, mainly propagating along the tunnel axis; while in a plastic surrounding rock section, the propagation speed decreases to 1500 meters per second, and there is a significant radial propagation component.

[0144] Extrapolating the predicted stress propagation direction and velocity after the stress state transition generates a distribution map of predicted arrival times of the stress wavefront. The stress wavefront refers to the spatial boundary where the stress value begins to change significantly during stress wave propagation. The distribution map of predicted arrival times is represented by isochrones, with each isochrone connecting spatial points where the stress wave is expected to arrive simultaneously. This distribution map visually demonstrates the spatiotemporal pattern of stress wave propagation from the point of abrupt change outwards, providing a basis for subsequent optimization of monitoring resource allocation.

[0145] In tunnel support structures, stress waves usually propagate along the shortest path of the support structure, but they are also affected by factors such as the non-uniformity of the support structure and the location of joints. By using the spatiotemporal evolution trajectory model of the stress field, the propagation law of stress waves under specific environments can be accurately predicted. For example, at the joint of the secondary lining of the tunnel, the propagation speed of stress waves changes abruptly, and the propagation direction also deflects. These characteristics are captured by the model and used for prediction.

[0146] The predicted arrival time distribution map of the stress wave is superimposed and matched with the spatial position of the encapsulation shell. The encapsulation shell is a protective shell for wireless sensors installed on the tunnel support structure. It integrates components such as stress sensors, data processing units, and wireless communication modules. By superimposing and matching, it is determined when the stress wave will arrive at each encapsulation shell position, so as to achieve accurate tracking of stress wave propagation.

[0147] The pre-wake-up monitoring unit is identified as a package whose predicted arrival time is earlier than the current time plus a preset advance. The preset advance is determined based on factors such as the monitoring system startup time and data processing delay, and is usually set to several seconds to several minutes. For example, if the current time is 10:00 and the preset advance is 5 minutes, then the package whose predicted arrival time is before 10:05 will be identified as the pre-wake-up monitoring unit.

[0148] The device identifier of the pre-wake-up monitoring unit and the difference between the predicted arrival time are encoded into a collaborative monitoring trigger command. The device identifier is a unique identification code for each package housing, and the difference between the predicted arrival time and the current time is the time difference between the predicted arrival time and the current time. The collaborative monitoring trigger command adopts a compact binary encoding format, which includes fields such as operation code, device identifier code, and time difference code, so as to facilitate wireless transmission and be correctly parsed by the monitoring unit.

[0149] The collaborative monitoring trigger command carries the wake-up countdown parameters of each pre-wake-up monitoring unit. The wake-up countdown parameters are calculated based on the difference between the predicted arrival times, taking into account the time required for the device to transition from a sleep state to a working state. For example, if the predicted arrival time of a stress wave for a certain package is 8 minutes after the current time, and the device needs 30 seconds to wake up, then the wake-up countdown parameter is set to 7 minutes and 30 seconds to ensure that the device has completed wake-up and is in normal working state before the stress wave arrives.

[0150] In practical applications, taking a high-speed railway tunnel as an example, when the monitoring system detects stress abrupt changes caused by geological activity at the entrance section, the above method is applied to process the stress abrupt changes. It is identified that the stress abrupt changes occur at the right arch shoulder of the tunnel entrance, and the abrupt change amplitude exceeds the threshold of 40%. Based on historical stress evolution data, a spatiotemporal evolution trajectory model is constructed to predict that the stress wave will mainly propagate along the tunnel axis at a speed of 2500 meters per second, while radiating towards the arch crown. A distribution map of the predicted arrival time of the stress wave is generated, showing that the stress wave will reach the monitoring point 50 meters away from the abrupt change point in 30 seconds, and the monitoring point 100 meters away in 60 seconds.

[0151] The predicted arrival time distribution map is matched with the positions of 60 encapsulated shells installed on the tunnel support structure to identify 15 pre-wake-up monitoring units. Cooperative monitoring trigger instructions are generated by encoding these instructions, which include the device identifiers and wake-up countdown parameters of these 15 units. The instructions are sent to the relevant monitoring units via a wireless network. These units wake up in advance according to the countdown parameters in the instructions, preparing to capture the stress wave information that is about to arrive.

[0152] This method enables the wireless stress monitoring system for tunnel support structures to accurately track stress wave propagation under limited energy conditions, providing technical support for safety status assessment and early warning of tunnel support structures. Especially in tunnel projects in high-risk areas such as earthquake-prone areas and fault zones, this method can significantly improve monitoring efficiency and early warning accuracy, thus ensuring project safety.

[0153] The present invention provides a wireless stress monitoring system for tunnel support structures, comprising:

[0154] The first unit is used to acquire stress response signals through an encapsulation shell installed on the surface of the tunnel support structure. The encapsulation shell integrates a stress sensor, a microprocessor, a battery, and a wireless communication module.

[0155] The second unit is used to perform time-frequency domain decomposition on the stress response signal on the microprocessor to extract stress features characterizing the stress state of the support structure.

[0156] The third unit is used to periodically upload the stress characteristics to the monitoring center through a wireless communication module, construct a spatiotemporal correlation diagram based on the stress characteristics, the spatiotemporal correlation diagram represents the stress evolution synchronicity and spatial gradient relationship at the location of each encapsulation shell node, and identify the dominant stress transmission channel based on the spatiotemporal correlation diagram to obtain the stress transmission channel topology.

[0157] The fourth unit is used to determine the stress convergence nodes of the support structure and construct a local stress transmission coordinate system based on the stress transmission channel topology. Combined with adjacent encapsulation shell nodes, it is used to deduce the stress distribution state around the stress convergence nodes and obtain the stress field of the high-risk area.

[0158] The fifth unit is used to identify the stress state transition time of the stress field in the high-risk area, construct an event-driven wake-up mechanism for the encapsulation shell, calculate the collaborative monitoring trigger command, and broadcast the collaborative monitoring trigger command to the encapsulation shells around the stress convergence node through the wireless communication module. The awakened encapsulation shells enter the synchronous sampling mode and transmit the sampling data to the monitoring center in real time.

[0159] A third aspect of the present invention provides an electronic device, comprising:

[0160] processor;

[0161] Memory used to store processor-executable instructions;

[0162] The processor is configured to invoke instructions stored in the memory to execute the aforementioned method.

[0163] A fourth aspect of the present invention provides a computer-readable storage medium having stored thereon computer program instructions that, when executed by a processor, implement the aforementioned method.

[0164] This invention can be a method, apparatus, system, and / or computer program product. The computer program product may include a computer-readable storage medium having computer-readable program instructions loaded thereon for performing various aspects of the invention.

[0165] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A wireless monitoring method for stress in tunnel support structures, characterized in that, include: Stress response signals are obtained by an encapsulation housing installed on the surface of the tunnel support structure. The encapsulation housing integrates a stress sensor, a microprocessor, a battery, and a wireless communication module. On the microprocessor, the stress response signal is decomposed in the time and frequency domain to extract stress features characterizing the stress state of the support structure; The stress characteristics are periodically uploaded to the monitoring center via a wireless communication module. Based on the stress characteristics, a spatiotemporal correlation diagram is constructed. The spatiotemporal correlation diagram represents the stress evolution synchronicity and spatial gradient relationship at the location of each encapsulation shell node. Based on the spatiotemporal correlation diagram, the dominant stress transmission channels are identified, and the stress transmission channel topology is obtained. Based on the stress transmission channel topology, the stress convergence nodes of the support structure are determined and a local stress transmission coordinate system is constructed. Combined with adjacent encapsulation shell nodes, the stress distribution around the stress convergence nodes is deduced to obtain the stress field in the high-risk area. The stress state transition moment of the stress field in the high-risk area is identified, an event-driven wake-up mechanism for the encapsulation shell is constructed, a collaborative monitoring trigger command is calculated, and the collaborative monitoring trigger command is broadcast to the encapsulation shells around the stress convergence node through the wireless communication module. The awakened encapsulation shells enter the synchronous sampling mode and transmit the sampling data to the monitoring center in real time.

2. The method according to claim 1, characterized in that, The stress response signal is decomposed in the time and frequency domain to extract stress features characterizing the stress state of the support structure, including: The stress response signal is segmented into sliding windows, and the peak value, mean, and variance of the stress response signal are calculated within each time window to construct a time-domain statistical feature sequence. The time-domain statistical feature sequence is subjected to a difference operation to calculate the peak increment, mean increment and variance increment between adjacent time windows to obtain the time-varying gradient feature. The time-varying gradient feature characterizes the dynamic change rate of the stress state of the support structure. The ratio of the peak increment to the variance increment is calculated to obtain the stress fluctuation concentration index. The stress response signal is subjected to frequency domain transformation to extract the main frequency and the corresponding spectral amplitude, and a frequency domain feature vector is constructed. The stress fluctuation concentration index is concatenated with the frequency domain feature vector to form a time-frequency fusion feature. The time-frequency fusion feature is subjected to principal component projection transformation, and the determinant value of the covariance matrix of multiple principal components is calculated to obtain the stress state stability quantification index. The determinant value characterizes the compactness of the distribution of the time-frequency fusion feature in the principal component space. The stress fluctuation concentration index, the time-frequency fusion feature and the stress state stability quantification index are combined as the stress feature output.

3. The method according to claim 1, characterized in that, Based on the stress characteristics, a spatiotemporal correlation diagram is constructed. This diagram characterizes the synchronicity of stress evolution and the spatial gradient relationship at the locations of each encapsulation shell node, including: The stress characteristic sequence of each package shell is periodically uploaded during continuous monitoring. The stress fluctuation concentration index is extracted and arranged in time order to form fluctuation time series data. The time-frequency fusion feature is extracted and arranged in time order to form time-frequency time series data. The fluctuation time series data is reconstructed in phase space and mapped to a high-dimensional phase space using a time delay embedding method. Phase point trajectories are extracted in the high-dimensional phase space, and the maximum Lyapunov exponent of the phase point trajectories is calculated. The maximum Lyapunov exponents of any two encapsulation shells are extracted to calculate the peak value of the cross-correlation function, which is used as the dynamic evolution synchronization degree and an evolution synchronization correlation matrix is ​​constructed. Singular value decomposition is performed on the time-frequency time-series data to extract the dominant singular value and calculate the time rate of change of the dominant singular value. The time rate of change difference between any two packaging shells is calculated. The spatial coordinates of each packaging shell are obtained and the spatial distance between the packaging shells is calculated. The time rate of change difference is divided by the spatial distance to obtain the spatial stress evolution gradient value and construct the spatial gradient correlation matrix. Based on the quantification index of the stress state stability of each encapsulation shell, the evolution synchronization correlation matrix and the spatial gradient correlation matrix are constructed into a spatiotemporal coupling correlation matrix. The encapsulation shell is used as a node, and the element values ​​of the spatiotemporal coupling correlation matrix are used as edge weights to construct a spatiotemporal correlation graph.

4. The method according to claim 1, characterized in that, Based on the spatiotemporal correlation diagram, the dominant stress transmission channels are identified, and the stress transmission channel topology is obtained, including: The edge weights and spatial distances between adjacent encapsulation shell nodes are extracted from the spatiotemporal correlation graph. The stress correlation intensity per unit distance is calculated. The spatial stress evolution gradient values ​​between encapsulation shell nodes are extracted and vectorized in three-dimensional space to obtain the dominant direction representing the preferred stress transmission direction. Based on the stress correlation intensity per unit distance and the dominant direction, multi-step diffusion with directional constraints is performed in the spatiotemporal correlation diagram. The product of the stress correlation intensity per unit distance on each side of the stress transmission path that meets the directional angle requirements is calculated to obtain the path stress transmission efficiency under directional constraints. The absolute difference of the stress fluctuation concentration index between adjacent packaging shell nodes on the stress transmission path is calculated and accumulated to obtain the path stress fluctuation attenuation index. Combined with the path stress transmission efficiency value, the comprehensive transmission performance index of the path is calculated, and multiple stress transmission advantage channels are selected. Extract the node identifiers, connection relationships, stress correlation intensity per unit distance, and comprehensive transmission performance indicators of the dominant direction and path of the stress transmission dominant channel, and output the stress transmission channel topology.

5. The method according to claim 1, characterized in that, Based on the stress transmission channel topology, stress convergence nodes of the support structure are determined and a local stress transmission coordinate system is constructed. Combined with adjacent encapsulation shell nodes, the stress distribution around the stress convergence nodes is deduced, resulting in the stress field of the high-risk area, including: Based on the stress transmission channel topology, the number of transmission paths entering each package shell node is counted, the variance of the angle between the dominant direction vectors of the entering paths is calculated to obtain the directional dispersion, and the product of the number of transmission paths entering is performed to obtain the stress convergence index. Package shell nodes with stress convergence index greater than the convergence threshold are regarded as stress convergence nodes. The neighboring nodes of the stress convergence node are extracted, a local stress transfer coordinate system with the stress convergence node as the origin is established based on the dominant direction, the radial distance of the neighboring nodes is calculated, the stress correlation intensity per unit distance of the neighboring nodes is vector-projected along the dominant direction to obtain the effective stress transfer component, the stress attenuation gradient is obtained by quotienting it with the radial distance, and the stress attenuation field function is obtained by spatial interpolation fitting. Based on the stress attenuation field function, a spatial monitoring grid is divided around the stress convergence node. The position parameters of each grid point in the local stress transfer coordinate system are calculated. Combined with the stress attenuation gradient of the neighboring nodes, the stress inference value of the grid point is obtained. Grid points whose stress inference value exceeds the stress risk threshold are extracted and connected and marked as high-risk area stress fields.

6. The method according to claim 5, characterized in that, Based on the stress attenuation field function, a spatial monitoring grid is divided around the stress convergence node. The position parameters of each grid point in the local stress transfer coordinate system are calculated. Combined with the stress attenuation gradient of neighboring nodes, the estimated stress values ​​of the grid points are obtained, including: Centered on the origin of the stress convergence node in the local stress transfer coordinate system, the outer boundary size of the spatial monitoring grid is determined according to the effective range of the stress attenuation field function. The spatial grid is then divided into equal-spaced sections along the three coordinate axes of the local stress transfer coordinate system to generate a set of grid nodes for the spatial monitoring grid and record their position parameters. For each grid node, a subset of neighboring nodes whose spatial distance is less than the influence radius is extracted. The stress attenuation gradient of each neighboring node is multiplied by the inverse of its spatial distance to the grid node to obtain the distance-weighted stress attenuation gradient. The distance-weighted stress attenuation gradients of all neighboring nodes in the subset of neighboring nodes are summed and normalized to obtain the comprehensive stress attenuation gradient of the grid node. Substitute the position parameters of the grid node into the stress attenuation field function to calculate the stress attenuation field function value at the grid node. Perform an inner product operation on the stress attenuation field function value and the comprehensive stress attenuation gradient in the dominant direction of the local stress transfer coordinate system, and superimpose the reference stress value at the stress convergence node to obtain the stress extrapolation value of the grid node.

7. The method according to claim 1, characterized in that, Identify the stress state transition moments of the stress field in the high-risk area, construct an event-driven wake-up mechanism for the encapsulation shell, and calculate the collaborative monitoring trigger command, including: For the stress projection values ​​of each grid point in the stress field of the high-risk area, statistical features are extracted within a sliding window. The deviation between the current stress projection value and the historical average within the sliding window is calculated. When the deviation exceeds the abrupt change detection threshold, the current moment is marked as the stress state transition moment, and the spatial coordinates of the grid point where the abrupt change occurs are recorded. Based on the stress inference values ​​of multiple historical moments before the stress state transition moment, a stress field spatiotemporal evolution trajectory model is constructed. The stress field spatiotemporal evolution trajectory model extrapolates and predicts the stress propagation direction and propagation speed after the stress state transition by fitting the motion trajectory of the stress inference values ​​in the time axis and spatial coordinate system, and generates a distribution map of the predicted arrival time of the stress wavefront. The predicted arrival time distribution map of the stress wave front is superimposed and matched with the spatial position of the packaging shell. The packaging shell whose predicted arrival time is earlier than the current time plus a preset advance is identified as a pre-wake-up monitoring unit. The device identifier of the pre-wake-up monitoring unit and the difference between the predicted arrival time are encoded into a collaborative monitoring trigger command. The collaborative monitoring trigger command carries the wake-up countdown parameters of each pre-wake-up monitoring unit.

8. A wireless stress monitoring system for tunnel support structures, used to implement the method as described in any one of claims 1-7, characterized in that, include: The first unit is used to acquire stress response signals through an encapsulation shell installed on the surface of the tunnel support structure. The encapsulation shell integrates a stress sensor, a microprocessor, a battery, and a wireless communication module. The second unit is used to perform time-frequency domain decomposition on the stress response signal on the microprocessor to extract stress features characterizing the stress state of the support structure. The third unit is used to periodically upload the stress characteristics to the monitoring center through a wireless communication module, construct a spatiotemporal correlation diagram based on the stress characteristics, the spatiotemporal correlation diagram represents the stress evolution synchronicity and spatial gradient relationship at the location of each encapsulation shell node, and identify the dominant stress transmission channel based on the spatiotemporal correlation diagram to obtain the stress transmission channel topology. The fourth unit is used to determine the stress convergence nodes of the support structure and construct a local stress transmission coordinate system based on the stress transmission channel topology. Combined with adjacent encapsulation shell nodes, it is used to deduce the stress distribution state around the stress convergence nodes and obtain the stress field of the high-risk area. The fifth unit is used to identify the stress state transition time of the stress field in the high-risk area, construct an event-driven wake-up mechanism for the encapsulation shell, calculate the collaborative monitoring trigger command, and broadcast the collaborative monitoring trigger command to the encapsulation shells around the stress convergence node through the wireless communication module. The awakened encapsulation shells enter the synchronous sampling mode and transmit the sampling data to the monitoring center in real time.

9. An electronic device, characterized in that, include: processor; Memory used to store processor-executable instructions; The processor is configured to invoke instructions stored in the memory to execute the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having computer program instructions stored thereon, characterized in that, When the computer program instructions are executed by the processor, they implement the method described in any one of claims 1 to 7.

Images